# Alin Boatca: full content index > Technical SEO, web development and digital PR from one technical lead. Diagnose the real root cause, build the fix, prove it on the live site. Practitioner: Alin Boatca. Based: Remote, United Kingdom. Contact: info@alinboatca.com. Site: https://alinboatca.com. Short index: https://alinboatca.com/llms.txt. --- # Overview The front door and how to get in touch. ## Home URL: https://alinboatca.com/ Updated: 2026-09-07 Alin Boatca is a technical SEO, web development and digital PR practice run by the person who does the work. The focus is finding the actual root cause of a ranking or performance problem, building the fix, and proving the result on the live site, not in a slide deck. ## Case studies URL: https://alinboatca.com/work/ Updated: 2026-09-07 Case studies covering a faceted-search build, edge security against a wp-login brute force, multi-layer cache invalidation, an internal-linking rebuild and a schema entity graph. Each is written up with the diagnosis and the measured outcome, including the wrong first theories. ## Rates URL: https://alinboatca.com/rates/ Updated: 2026-07-23 Transparent, published pricing. A standard rate card, a blended hourly rate for mixed work, and higher rates for adult and restricted sectors, each shown with its reason and none left to a private quote. ## Find a service URL: https://alinboatca.com/find-a-service/ Updated: 2026-07-08 A filterable directory of every SEO, web development and digital PR service on the site, built so a visitor can skip the long menu and narrow the whole catalogue by four facets: service discipline (technical, on-page, links and PR, audits, AI, WordPress and build, CRO, strategy, migration, industry and local), industry or sector, platform (WordPress, Shopify, agnostic) and location (UK-wide and the individual cities). A compact filter bar holds a text search and a dropdown for each facet, with live result counts on every option, shareable filtered URLs, and no form or data collection. It is the same faceted-search pattern built for client directories, applied to the practice’s own services. ## CV URL: https://alinboatca.com/cv/ Updated: 2026-09-07 A client-facing CV for the practice, published on the domain and not as a PDF, on the principle that every claim should be verifiable from a page on the same site. Covers the at-a-glance facts (discipline, base, fourteen years including regulated and restricted sectors, engagement model, the published blended rate, and that no account manager sits between the client and the practitioner), the selected work with the measured outcome and a link to the full write-up of each, and the working stack grouped by the job it does: search and analysis, performance, platform and build, infrastructure and security, and structured data. Also lists the seven operating standards: evidence over assumption, with every fix verified live before and after; root cause over symptom; a tight blast radius, with a backup first and targeted purges; SEO risk flagged when a change is proposed, before it ships; billing in the open, with time logged as it is spent and one published rate card; a living knowledge base on every project; and protecting the client, from a verified rollback path to declining paid link and PBN schemes outright. Closes on the site itself as the portfolio piece and offers a print view. ## Privacy notice URL: https://alinboatca.com/privacy/ Updated: 2026-06-30 A plain privacy notice. The site runs no analytics, sets no tracking or advertising cookies, and embeds no social pixels; the only personal data it handles is the name, email and message submitted through the contact form, used solely to reply. Processors are limited to Cloudflare (hosting, the form function and the Turnstile anti-spam check) and Resend (delivering the submission as an email). States the UK GDPR legal basis (legitimate interest in answering an enquiry), that data is never sold or used for marketing, retention (kept only as long as the enquiry is useful, deleted on request), that only an essential Cloudflare security cookie may be set, and the rights of access, correction and deletion by emailing directly, with the ICO named as the regulator. ## Clients outside the UK URL: https://alinboatca.com/international-clients/ Updated: 2026-07-10 What changes for a client outside the UK, and by how much. Publishes hours of genuine overlap against a UK working day by zone: Ireland full, Western Europe almost full, US Eastern about four hours, US Central about three, US Mountain about two, US Pacific about one, and Australia effectively none. Notes that the UK and US do not change their clocks on the same dates, so every figure shifts for a few weeks twice a year, which is why standing calls are set against the client local time. On money: invoices are issued in pounds at the same published blended rate with no overseas loading and no separate international price list, the client bank handles conversion so they pay the rate on the day, and tax treatment is left to their accountant. Sets out what does not change, including the same person diagnosing and implementing and the open ledger, and what does, including decisions moving to writing, deploys scheduled against the client traffic, and rankings checked from the target market instead of a UK browser. Names the cases where somebody closer is the better buy: US Pacific with a need for daily contact, Australia and New Zealand, non English markets which need native keyword research and copy, procurement requiring a local entity, and anyone who wants somebody in the room. ## Contact URL: https://alinboatca.com/contact/ Updated: 2026-05-02 Direct contact with the person doing the work, no account manager in the middle. A free 10-minute call is a diagnosis, not a pitch, and email replies usually come the same day. --- # SEO services Search work, from the whole-site view down to a single page. ## SEO services URL: https://alinboatca.com/seo-services/ Updated: 2026-02-26 The full menu of SEO work and how the individual services connect: technical, on-page, content, links and digital PR, with a note on which lever matters for a given site. ## Technical SEO services URL: https://alinboatca.com/technical-seo-services/ Updated: 2026-06-29 Ongoing technical SEO where the fixes are implemented, not just listed. Crawling and indexation, structured data, site architecture, rendering and Core Web Vitals, done on the codebase itself. ## On-page SEO URL: https://alinboatca.com/on-page-seo/ Updated: 2026-07-08 On-page optimisation for pages that are technically fine but still will not rank: titles and headings, internal linking, content that matches intent, and the cannibalisation and canonical problems that hold a strong domain down without anyone noticing. ## AI SEO URL: https://alinboatca.com/ai-seo/ Updated: 2026-06-29 AI SEO and generative engine optimisation: entity clarity so a model does not confuse you with someone else, structured data that resolves as one graph, content written in self-contained quotable passages, generated llms.txt and llms-full.txt, and an AI crawler policy you chose yourself. Admits that most of this is technical SEO under a new name, and that no one can guarantee a citation in a non-deterministic answer. ## LLM SEO URL: https://alinboatca.com/llm-seo/ Updated: 2026-06-18 LLM SEO, also sold as AI search optimisation, is the work that decides whether a large language model such as ChatGPT, Claude, Gemini, Perplexity or Copilot uses your site as a source when it answers. Explains that it overlaps almost completely with AI SEO and generative engine optimisation, that the proliferation of acronyms is a marketing habit and not a real taxonomy, and disambiguates the four terms while cross-linking the AI SEO and GEO pages. The different parts for a model: answers stated in self-contained passages that can be lifted whole, entity consistency across the third-party sources models already trust, correct description over mere presence, and measurement by prompt sampling because there is no rank to track. Refuses citation guarantees on non-deterministic output, and ships llms.txt and llms-full.txt on this domain as a near-zero-cost bet you can fetch and inspect. ## Programmatic SEO URL: https://alinboatca.com/programmatic-seo/ Updated: 2026-06-18 Programmatic SEO: generating a large set of pages from a data source and one template. Spells out that the technique is neutral and the data behind it decides whether it works, that Google ignores thin near-duplicate pages and at scale reads them as a site-wide quality problem, and that the valuable work is subtractive: deciding which combinations earn a page and which should not exist. Covers the data model and quality audit, mapping which combinations have real search demand and distinct intent, template design with genuine per-page substance, indexation and crawl control so budget reaches the pages that earn, internal linking that concentrates authority, and post-launch monitoring and pruning. Says when not to generate at all: when the data has to be invented, when existing pages already struggle to index, or when the combinations have no measurable demand. A development project as much as an SEO one, billed at the £95 blended rate and scoped from the data, whatever the page count. ## International SEO URL: https://alinboatca.com/international-seo/ Updated: 2026-06-03 International SEO for sites selling into more than one country. Choosing between ccTLD, subdomain and subfolder on the real trade-offs, hreflang implemented as a complete reciprocal set that survives releases, confirming Googlebot can reach every locale without being redirected by IP, duplicate handling across near-identical language variants, native keyword research, and multi-region migrations treated as the migrations they are. ## SaaS SEO URL: https://alinboatca.com/saas-seo/ Updated: 2026-06-30 SEO built for software companies, where the person who signs up is not always the person who pays. Product-led content, comparison and integration pages, and technical work on app-shell rendering and programmatic pages. ## B2B SEO URL: https://alinboatca.com/b2b-seo-agency/ Updated: 2026-06-24 SEO for B2B companies with long, considered sales cycles, measured against pipeline. Vanity traffic is not pipeline. Bottom-of-funnel commercial pages, earned authority, and the technical foundation underneath. ## Local SEO URL: https://alinboatca.com/local-seo-services/ Updated: 2026-06-15 Local SEO for businesses that serve a defined area: the map pack, Google Business Profile, local landing pages and the citation and review signals that decide who ranks in a city. ## Travel SEO URL: https://alinboatca.com/travel-seo/ Updated: 2026-06-03 Travel and hospitality SEO for guide sites, hotels, directories and booking journeys. Travel content is image heavy, taxonomy heavy and seasonal, so it fails in three specific ways: a taxonomy that generates thousands of URLs with nothing to say, galleries that take seconds to settle, and seasonal pages published months late, once the season has started. Covers indexation at scale, the image pipeline, internal linking that concentrates authority, multi-region targeting and listing structured data. Backed by two public live builds, London Guide UK and Lamplit London. ## Healthcare SEO URL: https://alinboatca.com/healthcare-seo/ Updated: 2026-06-05 Healthcare SEO for clinics and providers. Health content is judged against a higher standard, so what decides it is whether qualified authorship is legible: named clinicians as real entities with credentials, a medical reviewer and review dates connected in schema. Also multi-location clinic architecture where sites differ by more than a postcode, condition and treatment page consolidation, keeping patient data out of URLs and third party scripts, indexation control across practitioner directories, and Core Web Vitals on pages carrying booking widgets. ## Dental SEO URL: https://alinboatca.com/dental-seo/ Updated: 2026-06-06 Dental SEO for practices and groups. Most enquiries are won in the local map pack, so profile completeness, citation consistency and genuine recent reviews outweigh content volume. Above that sit implants and orthodontics, considered purchases needing treatment pages that give price ranges, recovery and risk. Covers clinician authorship schema with GDC registration, multi-practice architecture on one domain, and booking flow performance. ## Automotive SEO URL: https://alinboatca.com/automotive-seo/ Updated: 2026-06-08 Automotive SEO for dealerships and motor trade sites. Individual vehicle pages live for weeks, so the durable rankings belong to make, model and body-style pages. Covers the inventory lifecycle and status codes (redirect sold stock to a genuine equivalent or return 410, never a stub at 200), faceted search control so crawl budget is not spent on filter permutations, Vehicle and Product schema with accurate availability, branch-level local signals, and image-heavy listing performance. ## Real estate SEO URL: https://alinboatca.com/real-estate-seo/ Updated: 2026-06-09 Real estate SEO for estate agents and property sites. Listing descriptions arrive from the same feed every competing agent receives and disappear when the property sells, so ranking individual stock is building on sand. The durable pages are areas, property types and branches. Covers area page architecture with more local substance than a swapped place name, listing lifecycle and status codes, feed duplication and canonical handling, faceted search control, and branch-level local signals. ## Ecommerce SEO URL: https://alinboatca.com/ecommerce-seo-services/ Updated: 2026-06-15 Ecommerce SEO focused on category and product templates, faceted navigation, indexation control across thousands of URLs, and the structured data that wins rich results for a store. ## Shopify SEO URL: https://alinboatca.com/shopify-seo/ Updated: 2026-07-11 Shopify SEO that works with the grain of the platform: theme and Liquid changes, app bloat, the duplicate-URL patterns Shopify creates, and collection pages that rank. ## SEO migration URL: https://alinboatca.com/seo-migration/ Updated: 2026-07-10 Migrations planned, implemented and verified so rankings survive the move: a redirect map built from every URL, canonical and internal-link continuity, content parity checks, and monitoring through the settling window. --- # Links, PR and audits Authority, and finding out what is wrong. ## Link building URL: https://alinboatca.com/link-building/ Updated: 2026-06-30 Earned links, none of them bought: digital PR, manual outreach and linkable assets, alongside backlink audits and disavow decisions. Includes saying so when authority is not the constraint. ## Digital PR URL: https://alinboatca.com/digital-pr/ Updated: 2026-06-03 Digital PR that earns coverage and authority links from real publications using data, stories and reactive commentary, built to move rankings instead of just collecting logos. ## SEO audit URL: https://alinboatca.com/seo-audit-services/ Updated: 2026-06-13 A full SEO audit across the technical layer, content and links, delivered as a prioritised list with the evidence behind every finding, so the first thing you fix is the thing that matters. ## Technical SEO audit URL: https://alinboatca.com/technical-seo-audit/ Updated: 2026-06-15 The narrow, deeper audit focused on the technical layer: crawl and indexation, canonicals and duplication, structured data, JavaScript rendering and Core Web Vitals, each finding ranked by impact. ## Penalty recovery URL: https://alinboatca.com/google-penalty-recovery/ Updated: 2026-07-12 Diagnosis and recovery for sites hit by a manual action or an algorithmic drop: link profile clean-up and disavow, reconsideration requests, and the content and technical work that rebuilds trust. ## Small business SEO URL: https://alinboatca.com/small-business-seo/ Updated: 2026-06-09 Small business SEO sold as hours at a published rate, with no bundled monthly package hiding what the buyer paid for. Covers finding the actual bottleneck before anything is spent, local visibility, the two or three pages that could realistically rank, speed measured on real devices, and the parts an owner can do themselves without being billed. Includes the cases where the right answer is to spend nothing: no search demand, a brand new site with nothing to diagnose, or a job that is one afternoon for whoever already has access. ## Financial services SEO URL: https://alinboatca.com/financial-services-seo/ Updated: 2026-06-10 SEO for financial services firms and fintech, held to the Your Money or Your Life standard Google applies to financial content. The most common problem is not technical: qualified people publishing anonymously, so the firm is indistinguishable from a content farm. Covers named authorship with qualifications and regulatory registration that resolve in schema, compliance aware page structure where risk and cost disclosure is designed in from the start, consolidation of years of overlapping pages, the gating decision, freshness cycles on rate driven pages, and keeping applicant data out of analytics. ## White label SEO URL: https://alinboatca.com/white-label-seo/ Updated: 2026-07-18 White label technical SEO for agencies and web studios that sell SEO without a full time specialist on payroll. Work arrives under the agency brand, in their template, with no client contact unless the agency arranges it. Same £95 blended rate as any other client, with no reseller markup. Covers audits, diagnosis the agency cannot reproduce, migrations, schema, Core Web Vitals and build work. Findings are written to be forwarded as they stand, because decoding is where most white label arrangements break. ## Enterprise SEO URL: https://alinboatca.com/enterprise-seo/ Updated: 2026-06-10 Enterprise SEO for large sites, where decisions are made at the template and inherited by every page it generates. Covers crawl efficiency established from logs, template level architecture and internal linking rules, indexation at scale, structured data that validates across every page a template produces, and release verification. Findings are ranked by revenue exposure and implementation cost, not by issue count, and written as specifications a platform team can ship. Works alongside in house teams, never in place of them. ## Law firm SEO URL: https://alinboatca.com/law-firm-seo/ Updated: 2026-06-10 SEO for law firms and solicitors. Legal search is expensive and heavily scrutinised, and both point the same way: firms win when their real, qualified people are visible and legible to a machine. Covers solicitor entities with admission dates and SRA registration connected in schema to the content they wrote, practice area architecture that follows what clients search and ignores department names, winnable terms ahead of contested head terms, one location page for each office and none for the towns around it, legal content freshness cycles, and enquiry flow confidentiality. ## SEO copywriting URL: https://alinboatca.com/seo-copywriting/ Updated: 2026-06-10 SEO copywriting and content remediation, written by a person with no generated text. Weighted towards fixing what exists before adding to it: rewriting pages already ranking on page two, merging thin pages competing for one query, commercial page copy that answers the search instead of describing the company, and helpful content remediation where deletion is a legitimate answer. Also briefs for in house writers. Billed below the blended rate at £50 to £70 an hour. Not the right fit for volume work. ## Organic SEO services URL: https://alinboatca.com/organic-seo-services/ Updated: 2026-06-10 Organic SEO meaning the position is earned at every layer, with nothing bought. Covers technical foundations where most early movement comes from, architecture decisions that determine what a site can rank for at all, content that answers the query, earned links through digital PR and manual outreach with no paid placements or private blog networks, structured data, and reporting against the pages that earn, not against terms chosen because they were winnable. Includes backlink audits and disavow calls where a site carries a bad history. ## UK SEO agency URL: https://alinboatca.com/seo-agency-uk/ Updated: 2026-07-06 A remote UK SEO practice working UK hours with UK clients and rates published in pounds. Covers what being in this market changes: a different competitive set to US results for the same query, UK spelling and vocabulary, regulatory context in finance and health, local search structure and citation sources specific to the UK, overlapping working hours, and invoicing in pounds under UK law. Also lists the cases where a single practitioner is the wrong choice: several parallel workstreams needing guaranteed cover, procurement requiring formal continuity, or large scale content production. ## London SEO URL: https://alinboatca.com/london-seo/ Updated: 2026-08-29 SEO for London businesses from a remote UK practice with no London address. What ties the practice to the city is client work: two public builds, Lamplit London and London Guide UK, plus a retained private dental practice in central London. Covers what density changes, namely that local results reorder over a few miles so a citywide ranking number means little and the map pack moves by neighbourhood. Includes area pages for the boroughs a business covers and none for the rest, profile prominence, deciding whether local matters at all for a nationally selling business, and the technical layer underneath. ## SEO Leeds URL: https://alinboatca.com/seo-leeds/ Updated: 2026-07-23 SEO for Leeds businesses from a remote UK practice with no Leeds office. Leeds carries the softest measured keyword difficulty of any major English city: seo leeds and seo agency leeds both at 2, seo company leeds at 0, on 500 to 1,000 searches a month each, against 54 for seo agency london. Covers local search where customers visit, the deliberate decision to skip local work for Leeds businesses selling nationally, and the technical layer. Backed by a UK directory whose Leeds city page holds position two for its main term against roughly 14,000 searches a month. ## SEO Birmingham URL: https://alinboatca.com/seo-birmingham/ Updated: 2026-06-12 SEO for Birmingham businesses from a remote UK practice with no Birmingham office. The Birmingham market splits unusually wide: seo birmingham at difficulty 24 on 1,700 a month is contested, while seo services birmingham at 1, seo company birmingham at 1, birmingham seo at 3 and seo expert birmingham at 2 sit largely unclaimed on 500 to 700 a month each. Covers claiming the variants first, West Midlands area pages only where a business works, and planning the head term as a project without promising it. Backed by a directory city page at position three against roughly 15,000 searches a month. ## SEO Liverpool URL: https://alinboatca.com/seo-liverpool/ Updated: 2026-07-03 SEO for Liverpool businesses from a remote UK practice with no Liverpool office. Liverpool has no established firm holding the city terms: seo liverpool at difficulty 5 on 1,000 searches a month and seo agency liverpool at 4 on 700, against 42 for the equivalent Bristol term in a similar sized city. Covers local search where customers visit, skipping it where they do not, Merseyside area pages only where served, and the technical layer. Backed by a directory city page at position six against roughly 12,000 searches a month. ## SEO Milton Keynes URL: https://alinboatca.com/seo-milton-keynes/ Updated: 2026-07-18 SEO for Milton Keynes businesses from a remote UK practice with no MK office. seo milton keynes takes 1,100 searches a month at a keyword difficulty of 0, the largest single city term found sitting completely open, with seo agency milton keynes at 500 and 0, seo company milton keynes at 300 and 3, seo services milton keynes at 250 and 0 and seo consultant milton keynes at 100 and 0, so roughly 2,250 a month with no incumbent worth the name against Manchester equivalents at 13 to 35 held by domain ratings in the seventies. Argues that the softness has a cause: MK is disproportionately a business to business town, so the agencies here sell national campaigns to national clients and have no reason to compete for their own postcode. Spends most of the page on the question this raises, whether an MK business should buy local SEO at all, and answers that logistics, distribution, warehousing, manufacturers selling to trade and B2B software firms should not, because their buyers never search the town and their money belongs in the technical layer and the national terms. Local is worth having for trades covering the villages, appointment based practices, anything with a front door and businesses whose catchment ends nearby. Covers the two things the grid layout changes: there is no single centre, so the local pack in Bletchley differs from Wolverton and a citywide position means little, and the road network makes a real service area reach Bedford, Northampton and Aylesbury, which argues for a few genuine area pages and against the page per village doorway pattern. ## SEO Brighton URL: https://alinboatca.com/seo-brighton/ Updated: 2026-07-18 SEO for Brighton businesses from a remote UK practice with no Brighton office. Brighton has more digital agencies per head than almost anywhere in the country, which ought to make its local market brutal, yet seo agency brighton at 400 a month, seo brighton at 250, seo company brighton at 200 and seo services brighton at 200 all measure at a keyword difficulty of 0. The explanation is that Brighton agencies sell to London and abroad, so their own city terms are worth almost nothing to them and nobody has claimed them; the density that looks like a warning is why the ground is clear. Also examines the most interesting figure on the page, a traffic potential anomaly: seo brighton draws only 250 searches a month while the top ranking page is estimated to receive around 2,100 visits, roughly eight times the term volume, because the city generates an unusually rich tail of related search including courses, jobs, meetups, freelancer listings and agency comparisons. That is the opposite of Newcastle, where a 600 volume term delivers about 30 visits, and it means the usual arithmetic understates this city. Notes that seo agency sussex at 250 and difficulty 13 and seo consultant brighton at 150 and 12 carry the only measurable defence, so the four zeros come first. Assumes an unusually literate buyer, since many readers here have run campaigns themselves, and concedes that a Brighton agency may suit a client better if they want a team or people to sit with. ## SEO Glasgow URL: https://alinboatca.com/seo-glasgow/ Updated: 2026-06-12 SEO for Glasgow businesses from a remote UK practice with no Scottish office. Glasgow shows the widest difficulty spread of any UK city measured: seo consultant glasgow at 43 and seo agency glasgow at 27 are defended by consultants with personal authority, while seo glasgow, the plainest phrasing, sits at 3 on 800 searches a month. Covers taking the undefended city term and leaving the incumbents their strongest ground, Scottish legal accuracy for property, legal and financial content, and Scotland wide architecture. Backed by a directory city page at position two against roughly 20,000 searches a month. ## SEO Edinburgh URL: https://alinboatca.com/seo-edinburgh/ Updated: 2026-06-12 SEO for Edinburgh businesses from a remote UK practice with no Edinburgh office. August does not resemble the rest of the year, and the recurring mistake is publishing festival and seasonal pages into the season when they cannot establish themselves in time. Covers durable seasonal URLs reused and refreshed every year, Scottish legal accuracy where content written for England is not merely less relevant but wrong, the plain city terms at difficulty 9, and capacity planning for a traffic spike that arrives in a fortnight. Backed by a directory city page at position three against roughly 12,000 searches a month. ## SEO Manchester URL: https://alinboatca.com/seo-manchester/ Updated: 2026-07-20 SEO for Manchester businesses from a remote UK practice with no Manchester office. Written as the counterweight to the other city pages: where Leeds, Sheffield and Liverpool are argued to be soft because their terms measure at difficulty 0 to 5, Manchester is not, and the page says so. Roughly 11,000 searches a month across nineteen variants. Explains why the measured difficulty of 13 on seo agency manchester flatters the market, by pulling the actual first page: four established agencies with purpose built location pages at domain ratings of 72, 52, 47 and 37, plus Clutch at 91 and Semrush at 92 taking a slot each. Then the two things that make it winnable anyway: positions six and seven held by domains rated 30 and 26 with no measurable links to those pages, which makes the mid table reachable on page quality alone, and the local variants nobody defended, with local seo services manchester at 450 a month and local seo manchester at 400 both sitting at difficulty 0. Sets out the sequence, measurement by neighbourhood since results reorder between the centre, Salford and the southern suburbs, and spells out when a Manchester agency is the better buy. ## SEO Sheffield URL: https://alinboatca.com/seo-sheffield/ Updated: 2026-07-23 SEO for Sheffield businesses from a remote UK practice with no Sheffield office. All four main terms measure at the bottom of the difficulty scale: seo agency sheffield at 0 on 1,000 searches a month, seo sheffield at 0 on 700, seo company sheffield at 0 on 500 and seo services sheffield at 1 on 400. England’s fourth largest city, roughly 2,600 searches a month of commercial intent, and effectively nobody defending any of it. Explains what a difficulty of zero does and does not mean, namely that the barrier is effort, with no accumulated authority to overcome. Invites the reader to falsify the whole argument in two minutes by searching the terms and counting how many results are genuine local agency pages and how many are directories. Carries the caveat that 2,600 searches a month is a useful stream of local intent but not a business plan on its own. ## SEO Nottingham URL: https://alinboatca.com/seo-nottingham/ Updated: 2026-07-02 SEO for Nottingham businesses from a remote UK practice with no Nottingham office. The market splits unusually cleanly: seo agency nottingham is the single biggest term across any of the soft UK city markets measured, at 1,200 searches a month, and the only one with a genuine defender at difficulty 8, while seo nottingham sits at 1 on 900, seo company nottingham at 1 on 600 and seo services nottingham at 2 on 450. Argues for taking the 1,950 a month of undefended variants before going for the prize, because all four terms want the same page, the easy ones start returning while the hard one climbs, and the authority earned is the same authority the hard term needs. Warns explicitly against the opposite move of building four thin pages, one per phrasing, which splits the intent and is the doorway pattern the practice refuses to build for clients. ## SEO Leicester URL: https://alinboatca.com/seo-leicester/ Updated: 2026-06-03 SEO for Leicester businesses from a remote UK practice with no Leicester office. Almost everywhere else the agency and company phrasings are the most defended because they are what buyers type, and the plain city term is softer because it catches research traffic that never converts. Leicester runs the other way: seo agency leicester at difficulty 0 on 700 a month and seo company leicester at 0 on 350, against the generic seo leicester at 3 on 900. Argues that the 1,650 a month across the three commercial terms is worth more to a business than the larger head term even before accounting for it being easier, because the commercial phrasings filter out the students, jobseekers and researchers the generic term collects. Notes the pattern is consistent with local firms optimising for the phrase they would type themselves over the one a customer types, and hands the reader a two search test to verify it. ## SEO Newcastle URL: https://alinboatca.com/seo-newcastle/ Updated: 2026-07-22 SEO for Newcastle upon Tyne businesses from a remote UK practice with no Tyneside office. The only UK market measured where the agency phrasing outsells the plain city term, at 900 searches a month against 600, with seo company newcastle at 250 and difficulty across all three between 0 and 3. Because buyers arrive having already decided to hire, the page meeting them has a different job: skipping the introduction and answering what you charge, what you have done and who does the work. Also covers the genuine technical quirk of the market, that seo newcastle draws 600 searches a month while the top ranking page is estimated to receive only about 30 visits, most likely because Newcastle upon Tyne, Newcastle under Lyme 200 miles away and Newcastle in New South Wales split the intent so no page consolidates it. The practical consequences are to state Newcastle upon Tyne explicitly in the page, title and structured data, to name the surrounding places served, and to budget against raw volume without applying an optimistic multiple. ## SEO Surrey URL: https://alinboatca.com/seo-surrey/ Updated: 2026-07-21 SEO for Surrey businesses from a remote UK practice. Unlike every other location page on the site, Surrey is a county of 1.2 million people spread across Guildford, Woking, Epsom, Camberley, Farnham, Reigate and Staines with no single centre, so a searcher in Farnham and a searcher in Caterham get entirely different local results for the same words. Explains the mechanic: the map pack is computed from the searcher’s own coordinates and a county has none, so there is no ranking in the map pack for Surrey at all, only in the pack of the town you are in. The county terms are therefore won in the organic results alone, which suits trades, professional services and anything sold at a distance, and suits a single shop needing footfall far less. Numbers measured at seo agency surrey 800 a month at difficulty 0, seo surrey 700 at 2, seo services surrey 300 at 0 and seo company surrey 250 at 0, unusually cheap for a county carrying some of the highest commercial search values outside London. Declines to build a page per town, which is the doorway pattern, in favour of one real county page plus town pages only where a business operates. ## Generative engine optimisation URL: https://alinboatca.com/generative-engine-optimisation/ Updated: 2026-06-28 Generative engine optimisation and answer engine optimisation for ChatGPT, Gemini, Perplexity, Copilot and AI Overviews. Makes the case that most of what is sold as GEO is technical SEO with a different noun on the invoice, and isolates the three things that differ: answers stated in self contained passages a model can lift, entity consistency across the third party sources models already trust, and measurement by prompt sampling because there is no rank to track. Refuses citation guarantees, since model outputs are non deterministic and change between versions. Notes that no major provider has confirmed reading llms.txt and that Google has said it does not, while publishing llms.txt and llms-full.txt on this domain anyway as a near zero cost bet you can fetch and inspect. ## UK SEO difficulty index URL: https://alinboatca.com/uk-seo-difficulty-index/ Updated: 2026-08-02 An original dataset measuring keyword difficulty, monthly search volume and traffic potential for the plain seo place term across 38 UK markets, all pulled from Ahrefs Keywords Explorer on the United Kingdom database on a single date so the rows are comparable. Median difficulty across the set is 1.5 out of 100, fifteen markets measure a flat zero, and only four sit above 30. Four findings are drawn out. City size does not predict difficulty: Cardiff measures 54 on 700 searches a month while London measures 1 on 3,200, because difficulty records who bothered to compete and says nothing about how large the prize is. The plain term and the commercial term are different markets, with seo london at 1 against seo agency london at 54, a 53 point gap between two phrasings a buyer might use interchangeably, while Leicester runs the opposite way. The difficulty score is misleading at the top of the range, demonstrated by pulling the live first page for seo agency manchester, which reads 13 yet is held by agencies at domain ratings of 72, 52, 47 and 37 plus Clutch at 91 and Semrush at 92, with the counterweight that positions six and seven belong to domains rated 30 and 26 with no measurable links. And search volume does not predict traffic, shown by Brighton at 250 searches delivering an estimated 2,100 visits against Newcastle at 600 searches delivering about 30, a seventy fold difference between two rows that look comparable in a volume column. States its method, its date, what it does not measure including the map pack and commercial value, why one row was excluded as untrustworthy data, and the conflict of interest in a firm that sells SEO publishing a table showing most UK markets are winnable. Published under a Creative Commons Attribution licence with a citation line. ## UK SEO statistics URL: https://alinboatca.com/uk-seo-statistics/ Updated: 2026-08-02 A statistics page built only from original measurements taken for this site, which is why it is short. Most SEO statistics pages collect a hundred numbers from studies by other people, each a link away from a source itself citing someone else, and a good share trace back to nothing checkable. This one carries eleven figures drawn from two published datasets, the UK SEO difficulty index and the link spam log, with every number computed from the underlying data arrays at build time, so nothing is typed by hand and a statistic quoted here cannot drift from the study it came from. Each claim is written to survive being copied out of context: the figure, the scope and the measurement date sit in the same sentence, so a writer lifting one line still has everything needed to attribute it. On local search it reports a median keyword difficulty of 1.5 across 38 UK markets, fifteen markets at a flat zero, only four above 30, London as the largest market at difficulty 1, and a Pearson correlation between market size and market difficulty, computed and not asserted, that shows no meaningful relationship. On link spam it reports 130 unsolicited referring domains arriving in six days from a zero base, every one measuring zero organic traffic, twenty carrying Domain Rating 50 or above which demonstrates that DR is manufacturable because it contains no traffic term, forty linking out to 300 or more domains, and ninety three sitting on a single cheap TLD. States its method, both measurement dates, what it does not measure including the map pack and commercial value, that the spam figures are a sample of one site over six days, and the conflict of interest in an SEO practice publishing figures showing UK markets are winnable. Carries a copyable citation line per dataset and points at the open data repository where the difficulty figures are published as CSV with re-pull instructions. ## Client feedback URL: https://alinboatca.com/client-feedback/ Updated: 2026-08-01 The public record of client feedback, kept separate from the feedback form itself, which stays noindex because an indexed form is a spam magnet. Quotes are verbatim: sentences are trimmed to the strongest ones and nothing is reworded, with the full versions kept on file. The single edit anywhere on the page is disclosed openly, an opening He changed to Alin at the client instruction because the pronoun lost its antecedent once the message was cut. Clients decide whether their feedback is published at all and the form defaults to no, so the page notes that it shows only what there was permission to show and is not a complete record of everything received. Names are withheld and each client is described by sector and location instead, matching the confidentiality that also governs the case studies on the work page. Carries no Review or AggregateRating markup, and explains why: self serving review markup on your own organisation is a documented enforcement target, and removing exactly that markup is how a Search Console entity issue was closed on a client site, so adding it here to win a star in the results would contradict both the blog post and the case study. ## Link spam log URL: https://alinboatca.com/link-spam-log/ Updated: 2026-07-31 An original dataset recording every referring domain that appeared unrequested at a brand new site over six days, pulled from Ahrefs Site Explorer on a single date. The sample is unusually clean because the domain had zero referring domains beforehand, so nothing in the table can be confused with an existing link profile, and nothing was bought, solicited or submitted anywhere. One hundred and thirty domains arrived between 26 and 31 July, around twenty two a day, peaking at forty in one day. The central finding is that every single one measured zero organic traffic while twenty of them still carried a Domain Rating of fifty or above, which demonstrates in one table why Domain Rating alone is the metric easiest to manufacture: it is calculated from links and contains no traffic term, so a network that links densely to itself can lift its own score without ever being visited. The highest rated domain in the set measures DR 74 with no traffic and links out to 998 other domains, and the heaviest links out to 5,323, which makes outbound domain count the cheapest tell available and one almost nobody checks. Ninety three of the hundred and thirty sit on a single cheap TLD. The anchor text is vendor advertising copy naming the target domain. None of it is a real citation. Also explains, at length, why the disavow file was left empty: Google discards this class of link automatically, a careless disavow can remove links you earned, and the only real triggers are a manual action in Search Console or a paid link programme someone ran. Domain names are published as plain text, unlinked. Creative Commons Attribution with a citation line. ## Press and brand assets URL: https://alinboatca.com/press/ Updated: 2026-06-27 A press and brand assets page for anyone writing about or linking to the practice: the correct spelling of the name, a one line and a paragraph boilerplate that can be lifted and edited, downloadable SVG logo files as a monogram and a dark and light wordmark, the brand colour palette with hex values, and the quotable facts. An assets page and not a link to us page: it serves people who have already decided to mention the practice and never solicits a link with a badge and keyword rich anchor text, the scheme pattern the link building page declines. Includes usage notes asking that the marks are not stretched, recoloured or placed on busy backgrounds. ## Build roadmap URL: https://alinboatca.com/roadmap/ Updated: 2026-08-04 A public build state for this website, organised into three lanes: live, in progress and next. Every figure is measured on the live domain, so a visitor can verify any line from a browser. Live items include a static Astro build with no CMS, zero third party hosts on the homepage, custom XML sitemaps in which no two URLs share a lastmod, llms.txt and llms-full.txt generated from the same catalogue as the sitemaps, JSON-LD that parses and validates on every page, a self hosted Cloudflare Pages Function handling enquiries, four layers of spam and flood defence, SPF DKIM and DMARC at p=reject, six security header policies and an hreflang generator. The Next lane is published as well, so the unfinished work can be checked later alongside the finished work. ## Construction SEO URL: https://alinboatca.com/construction-seo/ Updated: 2026-06-18 SEO for construction companies and contractors. Built around the fact that the enquiry arrives months after the search, so measurement windows and content both have to match a long sales cycle. Covers separating commercial and domestic buyers, who search different phrases and are poorly served by a single services section, an image pipeline that keeps project galleries while fixing the speed they usually cost, project pages that state scope and value band and duration, area pages confined to where a contractor works, and accreditations such as CHAS and Gas Safe expressed as machine readable credentials instead of logo images. ## Insurance SEO URL: https://alinboatca.com/insurance-seo/ Updated: 2026-06-18 SEO for insurance brokers and firms. Starts from the position that comparison sites own the generic product terms and cannot realistically be displaced there, then targets the ground they handle badly: specialist and non standard risks, commercial lines with underwriting nuance, and questions where a broker judgement is the answer. Covers FCA aware content structure with disclosures designed into templates, YMYL authorship with named brokers and registration that resolves, publishing exclusions and price bands because premium figures date, quote funnel privacy, and freshness cycles on regulation dependent pages. ## SEO for accountants URL: https://alinboatca.com/accountant-seo/ Updated: 2026-06-18 SEO for accountancy practices. Two problems define the sector: self assessment demand peaks in January but pages published in December arrive too late to establish, and every practice describes the same services in the same order so nothing differentiates them. Covers durable seasonal URLs that carry over from one tax year to the next, specialism pages built around the clients a practice serves such as contractors inside IR35 or e-commerce sellers with overseas VAT, ICAEW and ACCA credentials marked up as data, local search where intent is local, rate and threshold freshness, and publishing fee bands. ## Plumber and trades SEO URL: https://alinboatca.com/plumber-seo/ Updated: 2026-06-19 SEO for plumbers, electricians and trades. Emergency search is resolved in the local pack before anyone reads a website, so profile completeness, accurate hours including out of hours, recent reviews and a tappable number outweigh any content programme. Covers service area business setup with the address hidden and coverage declared, prominence signals, job type pages for boiler replacement or rewires which differ in urgency and value, area coverage that matches where the jobs are instead of a page per town, and accreditation as structured data. Warns that an unanswered phone makes everything above it worthless. ## Restaurant SEO URL: https://alinboatca.com/restaurant-seo/ Updated: 2026-06-19 SEO for restaurants. The most expensive loss in the sector is a searcher who already chose you and lands on an aggregator page instead, so owning your own name is the first fight. Covers menus rebuilt as structured pages (a PDF menu is the most common technical fault in hospitality), opening hours including special and holiday hours because they drive open now filtering, an image pipeline for food photography which is the usual reason a restaurant site is slow, booking flows measured on a real phone, and multi site structure for groups. Argues most restaurants should not have a blog. ## SEO for architects URL: https://alinboatca.com/architect-seo/ Updated: 2026-06-21 SEO for architecture practices. Practice sites organise work by internal project name, which is meaningless as a search term, while clients search the thing they want and the constraint they fear: side return extension, listed building consent, barn conversion, green belt replacement dwelling. Covers project type architecture linking through to the schemes that demonstrate it, an image pipeline for portfolios that are consistently the heaviest sites measured, scheme write-ups giving brief and constraint and outcome, planning content that helps during a year long decision, and ARB and RIBA credentials expressed as machine readable data. ## Manufacturing SEO URL: https://alinboatca.com/manufacturing-seo/ Updated: 2026-06-24 SEO for manufacturers and industrial suppliers. Specification and part number queries are the highest intent and least contested searches in industrial buying, and most manufacturers lose them by keeping the answer inside a PDF or behind a search box a crawler cannot operate. Covers generating a crawlable page per product and part reference from existing data, specification markup instead of images of tables, catalogue architecture that avoids generating thin URLs at scale, multi market structure with hreflang and correct units, technical content aimed at specifiers, and the distributor channel conflict decided before it happens. ## Recruitment SEO URL: https://alinboatca.com/recruitment-seo/ Updated: 2026-06-26 SEO for recruitment agencies. Every filled role becomes a URL with nothing on it, so an agency site accumulates dead pages at the rate it fills jobs and teaches Google to crawl it less. Covers an expiry policy decided in advance and not left to a plugin producing soft 404s at scale, JobPosting structured data maintained with valid through dates so it never describes a filled role, crawl priority on category pages that persist over listings that will not, separate candidate and employer journeys since the employer side is where revenue sits and is usually neglected, and publishing real salary data as citable reference content. ## Adult SEO and directory builds URL: https://alinboatca.com/adult-seo/ Updated: 2026-07-23 Adult and restricted sector SEO plus custom site builds. The usual channels are closed here, so paid search is largely unavailable, mainstream press will not cover the sector, and the link networks that approach it are refused. What remains is technical SEO on heavy image dense listing sites, coverage earned through the sector trade press and adjacent non explicit wellness and lifestyle publishers, and custom builds. Escort agencies and directories are built from scratch to client preference: listing and profile structures, faceted search with real logic, moderation and verification tooling, reviews and galleries at scale, age assurance and labelling built into the templates, correct adult classification so search engines categorise the site and do not penalise it, and portable infrastructure so no single high risk host or payment processor can end the business. Billed at the published restricted rate of £220 an hour. Backed by an established anonymised UK directory ranking near the top nationally. ## Accessibility audit URL: https://alinboatca.com/accessibility-audit/ Updated: 2026-07-08 Website accessibility audits against WCAG 2.2 at level AA, the level almost every legal framework points at, including EN 301 549 in the EU. Built around the fact that automated scanning finds roughly a third of real issues: a tool reliably catches contrast ratios, missing alt attributes, unlabelled form inputs and absent landmarks, and cannot judge whether alt text describes the right thing, whether focus order makes sense, whether an error message tells anyone what to do, or whether a custom component announces itself correctly to a screen reader. Covers keyboard only journeys through navigation, forms and checkout, screen reader passes over the same journeys, structure and semantics, forms and error handling, media and motion including reduced motion, and auditing at the template level, where a broken component that appears on four hundred URLs gets fixed once. States the legal position without pretending to give legal advice: the Equality Act has long covered services delivered online, UK public sector bodies carry separate duties including an accurate published accessibility statement, and the European Accessibility Act has applied since June 2025, reaching UK businesses through who they sell to, wherever they are based. Warns against overlay widgets that promise compliance in a line of JavaScript, which cannot repair semantics they did not author and have been the subject of a large number of US lawsuits. Declines to offer a compliance certificate, legal advice, or testing with disabled users, which is a specialist service worth budgeting for separately. ## Conversion rate optimisation URL: https://alinboatca.com/conversion-rate-optimisation/ Updated: 2026-06-03 CRO for sites already getting meaningful traffic. Covers the technical causes of lost conversions that most CRO work ignores, page speed and layout shift on the pages that convert, form and checkout friction, and rigorous experiment design: a sample size worked out before the test runs, one change at a time, and calling a result only when the data supports it. Includes telling the client when a site has too little traffic for testing to be meaningful, in which case the answer is traffic first. ## Online reputation management URL: https://alinboatca.com/online-reputation-management/ Updated: 2026-07-10 Online reputation management built on the brand SERP, meaning what a search for your own name returns, and on neither of the two things most of the industry sells. States the refusals before the pitch: no fake or incentivised reviews, no review gating where only happy customers are asked, no networks of thin sites built to occupy page one, no paid placements dressed as editorial, and no promises to delete a result somebody else published. Explains that this is commercial advice as much as principle, since fake and incentivised reviews are explicitly unlawful under UK consumer protection law with the CMA able to act, and suppression networks are the same scheme pattern that earns manual actions. Separates the brand SERP into what you control, what you influence and what you do not, and is clear that genuine criticism belongs in the last category. Covers knowledge panel and entity accuracy, which is assembled from third party sources, correcting facts at source across the profiles and directories that feed everything downstream, ranking your own pages for your own brand terms, a review habit you can defend, and answering the question a bad result raises, since hiding that it was asked does not last. Treats what AI assistants say about a business as part of the same problem, fixable at the cited sources and out of reach of your own about page. Names when it is the wrong purchase: when the criticism is accurate and the product is the fix, when it is defamation or fraud and needs a solicitor, when nobody is searching for the brand at all, and when the client wants it resolved within a week. ## SEO consultant URL: https://alinboatca.com/seo-consultant/ Updated: 2026-07-03 Consulting for teams that need direction: strategy grounded in what competitors are doing, prioritisation, and a second opinion from someone who also does the implementation. --- # Build and WordPress Design, development and keeping a site fast and safe. ## Web development and design URL: https://alinboatca.com/web-design/ Updated: 2026-07-13 Web development services for sites that have to perform in search, not just look good: fast, accessible, well-structured front ends built so the SEO work has something solid to stand on. Includes ecommerce web design, treated as the different job it is, where the category template, not the homepage, carries most of the search traffic, layouts are designed against the real, untidy catalogue, checkout is measured on an actual mid range phone on mobile data, product schema is kept true to the live stock position, and the platform is chosen on catalogue size and who maintains it afterwards, whatever the builder prefers. Spells out that redesigning an existing store is a migration and should be treated as one. Separates Shopify development from Shopify SEO, which are sold as one thing and are not: development is theme work, Liquid templates, sections and blocks a client can use, metafields modelled properly so nothing is stuffed into a description, and removing the apps doing what a few lines of Liquid would do faster, since a store carrying eleven apps is usually carrying eleven render blocking scripts. Names the platform boundary at the outset, before anyone discovers it mid project: checkout is largely locked outside Plus and some URL structures cannot be changed at all. ## WordPress development URL: https://alinboatca.com/wordpress-development/ Updated: 2026-07-13 Custom WordPress development: bespoke themes and plugins, PHP work, and the careful build decisions that keep a site fast and maintainable instead of a pile of conflicting plugins. ## Website redesign URL: https://alinboatca.com/website-redesign/ Updated: 2026-05-20 A redesign treated as the migration it is: the look and structure change while redirects, canonicals, internal links and content parity protect the rankings the old site earned. ## Custom software and web apps URL: https://alinboatca.com/custom-tools/ Updated: 2026-07-13 Custom software development and bespoke web application development for when no off-the-shelf plugin fits, running from a single tool inside a site up to a full SaaS product: accounts and authentication, per-user data, dashboards and analytics, hosted separately from the marketing site. Also calculators, comparison and configurator tools, geolocation search, data pipelines, API integrations and the linkable assets that earn coverage. Public-facing tools are server rendered so they stay crawlable, and everything is documented for handover. ## WordPress SEO consulting URL: https://alinboatca.com/wordpress-seo-consultant/ Updated: 2026-05-13 WordPress-specific SEO: templates, taxonomies, plugin decisions and the rendering and pagination quirks that generic SEO advice tends to miss on a WordPress build. ## WordPress speed optimisation URL: https://alinboatca.com/wordpress-speed-optimisation/ Updated: 2026-06-03 Speed work measured on real field data, not a synthetic score: multi-layer caching, image and asset delivery, render-blocking resources and the Core Web Vitals that affect both users and rankings. ## Maintenance and support URL: https://alinboatca.com/wordpress-maintenance/ Updated: 2026-07-08 Ongoing maintenance done properly: updates tested before they reach the live site, security hardening, backups, uptime and change monitoring, handled by someone who knows the codebase. ## SERP preview and meta description length checker URL: https://alinboatca.com/tools/meta-description-length-checker/ Updated: 2026-07-16 A free browser based SERP preview tool that measures meta description and title tag width in pixels. Character counts mislead because Google truncates by pixel width, and a line of capitals is far wider than the same number of narrow letters. Shows a live approximate result preview with the text cut where it is likely to cut, on desktop and mobile budgets, alongside the measured width and the character count. Nothing is transmitted: the measurement runs in the browser using canvas text metrics. Lists its own limits, including that Google rewrites the majority of descriptions anyway, that the truncation boundary moves with device, locale and Google layout changes, and that an empty description is a legitimate choice on large long tail sites. ## Hreflang generator (free tool) URL: https://alinboatca.com/tools/hreflang-generator/ Updated: 2026-07-12 A free tool that generates a complete, reciprocal, self-referencing hreflang set from a list of locales and URLs, with x-default handled, output as HTML link tags, XML sitemap entries or an HTTP Link header. Flags the common faults as you type: invalid language or region codes, en-UK instead of en-GB, relative URLs and duplicate locales. Runs entirely in the browser, nothing is uploaded. ## Free tools URL: https://alinboatca.com/tools/ Updated: 2026-07-17 The index of the free browser based tools on this domain: a schema markup generator producing valid JSON-LD for the six types most sites need, a meta description and title checker that measures in pixels because pixels are how Google truncates, an hreflang generator that builds a complete reciprocal self-referencing set in three output formats, an llms.txt generator that builds both the short index and the full content version, and a robots.txt generator with presets for WordPress, Shopify, staging and AI crawlers. Every one runs entirely in the browser with nothing transmitted, which the page invites you to check for yourself in the network tab, and none of them requires an account, an email address or a trial. Explains why they are free: each began as something written for a client job, hosting a browser side tool costs effectively nothing, and putting it behind an email capture would amount to charging for the privilege of being marketed to. Also states what they are not, namely a substitute for judgement, since whether markup should exist at all or whether hreflang is even the problem are decisions no generator makes. ## llms.txt and llms-full.txt generator (free tool) URL: https://alinboatca.com/tools/llms-txt-generator/ Updated: 2026-07-13 A free browser based tool that generates a valid llms.txt file: an H1 name, a blockquote summary, optional context, and H2 sections each holding a curated markdown list of links, plus the skippable Optional section the proposed spec defines. Links are entered one per line as title, URL and an optional note, tolerating tabs so a spreadsheet paste works, and falling back sensibly when only a bare URL is given. Flags the faults as you type: a section with links but no heading, an entry with no URL, a relative address where an absolute one is required, and a missing summary. A toggle switches the same inputs between the two files. In full mode it builds llms-full.txt instead, giving every link its own content box and inlining what you paste under an H2 carrying the page URL, with any page left empty explicitly marked in the output so a half finished file cannot be shipped by accident. Content is keyed to the URL so it survives edits to the link list above. Output can be copied or downloaded, correctly named for whichever file is selected. Runs entirely in the browser, nothing is uploaded. Acknowledges the one real limit of the full version: the same origin policy stops a browser fetching your pages for you, so the content is paste or type, which is fine for a dozen pages and the wrong approach past that, where the file should be generated from whatever already holds the content. Unusually for a tool page it argues against overvaluing its own output: no major provider has confirmed reading llms.txt and Google has said it does not use it, so the case for it is that the file costs almost nothing and is worth having if the convention is adopted, not that it earns citations. Distinguishes llms.txt from llms-full.txt and from robots.txt, which solves the opposite problem, and points out that a model describing a business wrongly is almost always a third party source problem, which no file in the root directory fixes. ## robots.txt generator (free tool) URL: https://alinboatca.com/tools/robots-txt-generator/ Updated: 2026-07-15 A free browser based tool that builds a valid robots.txt from user-agent groups, allow and disallow rules and a sitemap line. Presets load real starting points: an open file, WordPress with admin-ajax.php allowed back in because blocking it breaks front end features on many themes, Shopify with cart, checkout, account and filter parameters excluded, a staging file that blocks everything, and a group per known AI training crawler. Rules are accepted as Disallow: /path, Allow: /path or a bare path treated as Disallow, and a spreadsheet style paste works. Warns on the faults that cause damage: a path with no leading slash, a sitemap given as a relative path when it must be an absolute URL, Crawl-delay which Google ignores, Disallow: / under a wildcard agent which blocks the entire site, and Disallow: / under Googlebot which removes the site from Search altogether. Unusually for a generator, the page argues against the assumption most users arrive with: Disallow controls crawling and has no say in indexing, so a blocked URL Google already knows about can still be listed with no description, blocking a URL guarantees a noindex tag on it is never read because reading it requires the fetch that was just refused, and the file is public so listing a private path advertises it. Separates Google-Extended, which governs AI training and does not affect Search rankings, from Googlebot, which does. States that robots.txt is honoured by convention with nothing to enforce it, so a determined scraper is a firewall problem instead. Runs entirely in the browser, nothing is uploaded. ## Schema markup generator (free tool) URL: https://alinboatca.com/tools/schema-markup-generator/ Updated: 2026-07-13 A free tool that generates valid JSON-LD structured data for the six types most sites need: Organization, LocalBusiness, Article and blog post, FAQPage, BreadcrumbList and Product. Fill the fields for a type and the tool builds the markup live, dropping empty fields, and flags the required ones that are missing. Carries notes on what each type earns as well as what validates: that FAQPage rich results have been restricted to government and health sites since 2023, that review stars need genuine rating data, never invented numbers, that a LocalBusiness address must match the Google Business Profile exactly, and that markup which contradicts the visible page is worse than none. Runs entirely in the browser, nothing is uploaded. --- # Guides Long write-ups from problems solved on live sites. ## Guides index URL: https://alinboatca.com/guides/ Updated: 2026-07-24 The guides section: in-depth write-ups drawn from problems diagnosed and fixed on live client sites, including the wrong turns. None of it is general SEO advice written for traffic. ## Every Search Console index status, explained URL: https://alinboatca.com/guides/search-console-statuses/ Updated: 2026-07-22 A pillar page for reading the Google Search Console Page Indexing report, on the principle that the terse status labels are not errors but literal descriptions of a decision Google made. Grouped by whether action is needed: usually fine, where Google is working correctly (alternate page with proper canonical tag, page with redirect); worth investigating, where the fix depends on the exact wording (discovered vs crawled currently not indexed, duplicate without user-selected canonical, soft 404); and the crawl-versus-index confusions (robots.txt does not stop indexing, and that Google discovers pages by following links, without needing them submitted). Ties together the individual status guides and posts, and frames the whole report as a series of decisions that are each either acceptable or changeable. ## A strong domain ranking for nothing URL: https://alinboatca.com/guides/strong-domain-ranking-for-nothing/ Updated: 2026-07-15 A DR-72 page ranked for zero keywords while DR-16-to-32 competitors held page one, so authority was clearly not the constraint. The four on-page causes that held it down: a canonical pointing at a redirect, two pages splitting the same intent, machine-spun copy, and a 25-link templated block spreading equity thin. No links were bought; the budget for a link campaign was never spent because the diagnosis proved it would not have worked. ## Migrate without losing rankings URL: https://alinboatca.com/guides/migrate-without-losing-rankings/ Updated: 2026-07-19 A migration checklist built from the failures that drop traffic. Capture the current state first, build a redirect map from every URL one-to-one with no chains, move redirects, canonicals and internal links together, confirm the staging noindex and robots rules are gone, check content parity on the pages that ranked, then watch Search Console and the server logs through the settling window. Most post-migration drops trace back to an incomplete redirect map, a canonical left aimed at the old site, or content that changed. ## Discovered, currently not indexed URL: https://alinboatca.com/guides/discovered-currently-not-indexed/ Updated: 2026-07-24 The Search Console status explained: Google knows the URL exists, usually from a sitemap or a link, and has not requested it. Nothing on the page caused it because nothing on the page has been read, which makes it the opposite of Crawled currently not indexed and means the fixes have almost nothing in common. Covers why Google rations crawl, the five common reasons it estimates a URL is not worth the request, and the uncomfortable one: on a new domain with no external links this is usually not a fault to fix at all. Lists what moves it, led by internal links from pages that already get crawled and by reducing URL sprawl, and what wastes time, including resubmitting the sitemap, bulk Request Indexing, the Indexing API outside job postings and livestreams, and editing lastmod dates to look fresh. ## Crawled, currently not indexed URL: https://alinboatca.com/guides/crawled-currently-not-indexed/ Updated: 2026-07-23 Google fetched the page, read it, and chose not to index it, which is a judgement about the content. Nothing technical is broken. Distinguishes it from "discovered, currently not indexed", which is a crawling problem with a different fix. Five causes in order: nothing on the page that is not already elsewhere, no internal links treating it as important, crawl allowance spent on parameter and filter URLs, locale or cross-domain duplication, and simply being new. Includes why requesting indexing repeatedly cannot help, and why the correct answer is often to delete the page instead of improving it. ## Duplicate without user-selected canonical URL: https://alinboatca.com/guides/duplicate-without-user-selected-canonical/ Updated: 2026-07-23 The opposite of the alternate page status: here no canonical was declared, so Google selected one from signals like internal links and URL tidiness, and it frequently picks a parameter URL or protocol variant over the page you would promote. Where the duplicates come from: tracking parameters, trailing slash and protocol variants, faceted navigation at scale, and near-identical locale pages. The fix in order is decide the owning URL, add self-referencing canonicals everywhere, redirect variants and not just canonicalise them, point internal links at the canonical, and use hreflang in place of canonical between locales. Includes why noindex is the wrong tool here. ## How to check whether AI assistants recommend your business URL: https://alinboatca.com/guides/check-if-ai-recommends-your-business/ Updated: 2026-07-23 A repeatable method for finding out how ChatGPT, Gemini, Perplexity, Copilot and AI Overviews describe a business, without a tool or a subscription. Covers swapping keywords for the questions customers ask, running each at least five times in fresh conversations while logged out, and recording four things each time: presence, the exact description used, who appears alongside you, and what was cited. Argues that reading the citations is the most actionable output, because the sources models trust are usually third party directories and publications, rarely your own site, so a wrong description has to be corrected at its source. Sorts findings into wrong, vague and absent, and says to fix them in that order. Sets out what the method cannot do: no volume figure exists, causation cannot be established, and the noise between two runs a week apart usually exceeds the signal. ## Soft 404 errors URL: https://alinboatca.com/guides/soft-404-errors/ Updated: 2026-07-23 A soft 404 is a mismatch between the HTTP status and the content: the header says success, the page says nothing is here. Five patterns cause it: a custom error template returning 200, empty search and filter results, discontinued products left as stubs, pages thin enough to read as empty, and content that only exists after JavaScript runs. Covers which status code each case should return, why 410 is the accurate response for content you removed and will not bring back, and why blanket-redirecting expired URLs to the homepage clears the report while throwing away the signal. ## Alternate page with proper canonical tag URL: https://alinboatca.com/guides/alternate-page-with-proper-canonical-tag/ Updated: 2026-07-23 This status normally means Google found a duplicate, read the canonical you declared, and honoured it, so no action is needed. The two minute check: follow the canonical, confirm it returns 200 without redirecting, confirm the destination is the page you wanted and is itself indexed. The four cases where it is hiding a problem: a canonical pointing at a redirect (found on a live DR 64 site that was ranking for nothing), a canonical aimed at a page that is not really equivalent, pagination canonicalised to page one, and faceted URLs consuming crawl budget at scale. ## Stopping a wp-login brute force at the Cloudflare edge URL: https://alinboatca.com/guides/wordpress-brute-force-cloudflare/ Updated: 2026-06-20 A wp-login page absorbing 46% of origin requests at 21,626 hits a day. How to check whether the attack is slowing the site, stop it at the Cloudflare edge in minutes with an admin allowlist and a rule on the login path, and avoid paying for a bigger server to absorb traffic that should never reach the origin. ## Why your WordPress cache will not clear URL: https://alinboatca.com/guides/wordpress-cache-invalidation/ Updated: 2026-07-19 Multi-layer cache invalidation across the CDN edge, OPcache, Redis and application transients, and the failure mode almost nobody checks: curated pages that query content without being taxonomy archives, so no term-based purge ever reaches them. Includes the object-meta hook trap where turning a flag off deletes the meta so neither the add nor the update hook fires. --- # Blog Shorter, opinionated takes on search. ## Common SEO myths, debunked URL: https://alinboatca.com/blog/seo-myths/ Updated: 2026-07-24 A pillar page collecting the SEO myths this blog debunks, on the observation that almost every one shares a shape: a proxy mistaken for the real thing, a number that becomes a target, or a genuine signal inflated into a lever. Grouped by where they do most damage: signals people overrate (Core Web Vitals, bounce rate, meta descriptions, keyword density), penalties that are not penalties (outranked not penalised, duplicate content, disavow), getting found and indexed (submitting to Google, the sandbox, robots.txt, page with redirect), links and third-party metrics (Domain Rating, nofollow), and content and strategy (a page per keyword, traffic as a goal, crawl-export audits, FAQPage schema, AI Overviews). Each links to the full argument, and the pattern to keep is asking of any rule where it came from and whether it is the real thing or a stand-in. ## Traffic is not the goal URL: https://alinboatca.com/blog/traffic-is-not-the-goal/ Updated: 2026-06-04 Traffic is the easiest number to grow and the easiest to mistake for progress, because a chart of sessions never shows who those sessions were or whether they wanted what you sell. Traffic and outcome can move in opposite directions: a site can double visitors and lose enquiries when the new visitors are the wrong ones. The wrong traffic is not free, it costs money to attract, load to serve, and it muddies conversion and behavioural data by averaging two populations. Chasing volume also pulls effort from the commercial pages where buyers land. Aim at the outcome, prioritise buying intent over raw volume, and convert the qualified visitors you already have, which matters most in long sales-cycle B2B measured against pipeline, not sessions. ## The Google sandbox is a myth, but new sites wait URL: https://alinboatca.com/blog/the-google-sandbox-is-a-myth/ Updated: 2026-04-09 The sandbox theory says Google applies a timed suppression to every new domain, a probation period no work can shortcut, which is comforting because it puts the cause outside your control. Google has rejected the fixed-timer claim, yet the waiting is real. The resolution is that a new domain arrives with almost none of the evidence Google ranks on: no links vouching for it, no history, unproven content, so it sensibly loses to established pages until it earns its place. That is correct behaviour, not a filter, and the competitiveness of the query sets how long it takes. "I am in the sandbox" says wait for a timer; "I have not earned signals yet" says go and earn them through links, coverage, internal structure and winnable terms first, while getting the technical foundation right. ## Keyword density is not a thing URL: https://alinboatca.com/blog/keyword-density-is-not-a-thing/ Updated: 2026-02-12 The two to three percent keyword density figure is a fossil from early search engines that really did lean on term frequency, an era that ended badly with keyword stuffing. Modern search works with meaning and intent, not a bag of words to count, so a page can rank for a term it uses once. When an audit tool grades density red the target came from the tool, not Google, which has never published an ideal density because none could be right across a disclaimer, a recipe and an essay. Making the ratio the brief means jamming the phrase in until the meter turns green, which stiffens the writing and drifts back toward stuffing. Read the page aloud instead: cover the subject the way someone who knows it would, and fix the on-page basics that carry weight. ## You do not need to submit your site to Google URL: https://alinboatca.com/blog/you-do-not-need-to-submit-your-site-to-google/ Updated: 2026-04-16 Google discovers pages by crawling links, so the moment one page it already crawls links to your site you are discoverable with nothing submitted, which is why the submit-to-500-engines services sell a solved problem and occasionally drop your domain into spammy neighbourhoods. The legitimate tools are precise in what they do: verifying in Search Console and submitting an XML sitemap helps discovery and reporting, URL Inspection nudges a page into the crawl queue, and none of it is ranking help or an indexing guarantee. When a page will not index the cause is almost never a missing submission but a quality or crawl-budget decision, covered by the discovered-currently-not-indexed status. The real levers are internal links, an earned external link, removing accidental blocks, and being useful. ## Nofollow links still have value URL: https://alinboatca.com/blog/nofollow-links-still-have-value/ Updated: 2026-03-19 Dismissing every nofollow link as worthless rests on an outdated model where a link did one thing, pass authority, and nofollow switched it off. Since 2019 nofollow, sponsored and ugc are hints Google may or may not use, so claiming a nofollow link does nothing asserts knowledge of an internal decision nobody outside Google has. More importantly a link was never only an authority pipe: it sends referral traffic that converts, builds brand and name searches, gets pages discovered, and sits in a profile humans read. A natural profile is a mix of followed, nofollow, ugc and sponsored, and the nofollow links are part of what makes it look earned; a profile of followed links alone looks engineered. The right unit of decision is the placement and its audience, not the attribute, though paid links must still be marked. ## Bounce rate is not a ranking factor URL: https://alinboatca.com/blog/bounce-rate-is-not-a-ranking-factor/ Updated: 2026-02-26 Bounce rate is calculated inside your own analytics from your own tracking, so Google Search does not have it and cannot rank on data it cannot read; Analytics and Search are separate products with a wall between them precisely so site-controlled scripts cannot drive rankings. The sophisticated version swaps in dwell time, but there is no simple per-page stopwatch to move, and time on page is worth caring about for the sake of the user, not the ranking. One bounce figure is ambiguous: a high rate can mean the page was terrible or that it answered the question completely, and a blog post, contact page and checkout should have different healthy rates. Measure what you want the page to do, treat bounce as a flag to investigate, never a verdict, and look for ranking causes on the page and in the results. ## Duplicate content is not a penalty URL: https://alinboatca.com/blog/duplicate-content-is-not-a-penalty/ Updated: 2026-03-05 The duplicate content penalty is a durable and damaging myth, because the fear leads people to noindex pages that should rank and delete content that was fine. Google has said more than once that there is no penalty for ordinary duplication: it clusters equivalent pages, picks one canonical to show, and sets the others aside, folding their signals in. The real cost is not punishment but inefficiency: without a declared canonical Google may index the wrong URL, and on large sites crawling near-identical parameter URLs wastes crawl budget. The fix is to declare the choice yourself with self-referencing canonicals, redirects for true variants, consistent internal links, and hreflang between languages, where a canonical is the wrong tool. ## You are not being penalised, you are being outranked URL: https://alinboatca.com/blog/you-are-not-penalised-you-are-outranked/ Updated: 2026-03-12 Penalty is the first word people reach for when rankings fall, because it puts the cause outside the site. A manual action is rare and announces itself in the Search Console Manual Actions report by name; if that report is clean, there is no manual action. The far more common causes are not penalties: a competitor improved on merit, an update rebalanced the results, or something broke on your own side such as a dropped redirect, a stripped internal link, a canonical pointing at the wrong URL, or a noindex escaped from staging. The label decides the investigation, so a wrong label hides the real problem: check Manual Actions, match the drop date against updates and your own deploys, then diagnose without a preferred conclusion. ## Meta descriptions do not affect rankings URL: https://alinboatca.com/blog/meta-descriptions-do-not-affect-rankings/ Updated: 2026-02-19 The meta description is not a ranking factor and Google rewrites the majority of snippets, swapping your generic line for one tailored to the query, so a whole category of audit busywork is misplaced. Its only influence is on whether a person who already sees your result clicks it, which makes it ad copy for a listing you did not pay for. Writing a bespoke description for every long-tail URL is effort with almost no return; spend it on the few commercial and head-term pages where the click is contested and valuable, and let Google generate the rest. Length is judged by pixel width not character count, which is why a pixel-based checker beats counting to 155. ## You do not need a blog post for every keyword URL: https://alinboatca.com/blog/you-do-not-need-a-blog-post-for-every-keyword/ Updated: 2026-05-21 Ten keywords around one intent do not need ten posts, they need one page. A keyword is a phrasing and an intent is what the person wants, and building a page per phrasing creates thin pages competing with each other for the same result, splitting the authority one page would concentrate. This is cannibalisation, and the fix is subtractive: merge pages sharing an intent into one and redirect the rest. Publishing volume feels like progress and is easy to sign off, while the harder and more valuable work is deciding what to cut, sometimes deleting a thin page to raise traffic. New pages are right only when the intent is uncovered. ## The disavow file is a loaded gun URL: https://alinboatca.com/blog/the-disavow-file-is-a-loaded-gun/ Updated: 2026-03-26 Disavowing links became routine hygiene, and it is not hygiene: it is a permanent instruction to Google to ignore links pointing at your site, at your own risk. The premise that spam links hurt you is a decade out of date, because Google ignores the vast majority of low quality links and does not penalise the target. Tool-generated toxicity scores are a vendor model, not Google judgement, and routinely flag good links, so feeding one into the disavow tool can demote the earned links that took years to build. The tool has one narrow correct use: a manual action for unnatural links, or a known history of bought links, handled as part of a deliberate recovery and never run on a schedule. ## Core Web Vitals are a tiebreaker URL: https://alinboatca.com/blog/core-web-vitals-are-a-tiebreaker/ Updated: 2026-05-07 Core Web Vitals are a genuine ranking signal and a small one that behaves like a tiebreaker: they help decide between comparable pages and do almost nothing for a page losing on relevance or authority. The score that counts is field data from real visitors, not a lab score a tool invents. Optimising every page when only the close contests need it wastes a fortnight shaving milliseconds off pages that are losing by a mile on substance. The better reason to fix speed is the visitor, not the ranking: layout stability and load time show up in completed forms and checkouts, and the same technical work that makes a page crawlable usually makes it fast. ## Most SEO audits are a crawl export URL: https://alinboatca.com/blog/most-seo-audits-are-a-crawl-export/ Updated: 2026-07-11 A crawler generates a two hundred page audit in seconds, so the length is no sign of thoroughness. It is the tell. Volume is not severity: two hundred thousand missing meta descriptions rarely matter while eleven canonicals pointing at a redirect chain usually do. The work of an audit is subtractive, deciding what to ignore and ranking the handful of problems that sit between what a site ranks for and what it is trying to rank for. How to tell a real audit from a crawl export before paying: ask what the first three items will be and why those three. ## AI Overviews are not a traffic apocalypse URL: https://alinboatca.com/blog/ai-overviews-are-not-a-traffic-apocalypse/ Updated: 2026-07-20 AI Overviews do not remove clicks from search in general. They remove them from a specific class of query, the short factual uncontested ones that featured snippets were already answering. The useful question is not whether they hurt but which pages exist only to answer a fact a machine can now answer completely. Decision, comparison and expertise pages are far less exposed and can gain from being cited. The response is mostly technical SEO done properly plus sampling how assistants describe you, since a generated answer has no rank to track. ## Domain Rating is a scoreboard, not a strategy URL: https://alinboatca.com/blog/domain-rating-is-not-a-google-metric/ Updated: 2026-04-02 Domain Rating is an Ahrefs estimate of backlink strength, not something Google publishes or uses. It is a fast way to read a situation and a bad thing to aim at, because a proxy that becomes the target gets gamed: the cheapest way to raise DR is to buy the high-DR links the number rewards, moving the figure without moving a ranking that matters. Earned authority and a bought metric can share the same number and look completely different in context. What to report instead: rankings and traffic on the pages that bring work, with DR as weather and not as a destination. ## Blog index URL: https://alinboatca.com/blog/ Updated: 2026-08-02 The blog: shorter, more opinionated pieces than the guides. Reactions to a Google change, a position held against the grain, or a best practice that stopped being one while nobody was looking. ## Page with redirect is not an error URL: https://alinboatca.com/blog/page-with-redirect-search-console/ Updated: 2026-04-30 The URL is not indexed because the destination is, which is the point of a redirect. It only causes alarm because it sits in the not indexed section of the report. Three cases need a closer look: redirect chains where each hop dilutes signal, mass redirects to a non-equivalent page such as the homepage (which Google may treat as a soft 404 anyway), and redirecting URLs still listed in sitemaps, navigation or ad landing pages. Includes the one-command check to count hops. ## Robots.txt does not stop indexing URL: https://alinboatca.com/blog/robots-txt-does-not-stop-indexing/ Updated: 2026-04-23 robots.txt controls crawling, noindex controls indexing, and people use the first when they mean the second. Google can index a URL it never fetched if enough pages link to it, which is why "indexed, though blocked by robots.txt" exists. The trap: adding a noindex to a page that is still blocked cannot work, because Google has to read the page to obey the instruction telling it not to. The order that works is remove the block, let the noindex be seen, wait for the drop, then restore the block if crawl budget matters. ## Your ranking is not a single number URL: https://alinboatca.com/blog/your-ranking-is-not-a-single-number/ Updated: 2026-01-08 There is no single result page any more: two people searching the same phrase at the same moment can be shown different results in a different order with different features between them. Location is the largest variable and often narrower than a city, which is why one citywide number for a multi-branch business is close to meaningless; device, signed-in history and the number of organic slots left above the fold all move it too. A rank tracker fixes those variables and reports one sample under one configuration, which is legitimate and useful for comparing change over time, but is not your position. Search Console average position is closer to the truth because it averages real impressions across real conditions, with the caveat that an average hides the spread: steady at eight and swinging between three and fifteen produce the same figure and are different situations. The single number invites optimising the measurement, picking flattering tracking locations and chasing head terms while the queries producing enquiries go unwatched. ## Google rewrites your title tags too URL: https://alinboatca.com/blog/google-rewrites-your-title-tags/ Updated: 2026-01-02 Google has generated result titles since 2021, drawing on the title tag, headings, anchor text and the page itself, on the reasoning that a tag is written once for every possible query while a result is shown for one specific query. A rewrite changes what is displayed, not how the page is assessed: the tag still counts as an on-page signal, so a rewrite is not a rejection and the response is never to stop writing them carefully. Common triggers are length, since a title long enough to truncate is often trimmed and the limit is pixel width, not character count; repetition and keyword chains; boilerplate that makes every page look the same so Google looks to the H1 instead; a mismatch where the H1 describes the page better, in which case the rewrite is feedback; and an empty or useless tag. Write the clearest description of the page, front-load the substance, check what is displayed for the queries in Search Console, and reserve real effort for the pages whose clicks are contested. ## Alt text is not a keyword slot URL: https://alinboatca.com/blog/alt-text-is-not-a-keyword-slot/ Updated: 2025-12-18 Alt text is the replacement for the image: read the sentence with the alt substituted for the picture and ask whether the reader learned what they needed. Search engines do read it, and it is how an image is understood well enough to appear in image search, but the version written for a person is reliably the version that works for the machine. Stuffing fails on every count, since a screen reader announces each image and reads its alt aloud, so a dozen images all described with the same keyword phrase tell the user nothing, and Google's image guidelines name keyword-stuffed alt text as spam. Write specifically about what the image shows, let surrounding context and captions carry their share, keep it to a sentence and put long descriptions such as chart data in the page content. Decorative images take an empty alt so screen readers skip them, which is the point most audits get backwards. On image-heavy sites, how the images are served usually matters more than the alt text. ## Exact-match domains are not a shortcut URL: https://alinboatca.com/blog/exact-match-domains-are-not-a-shortcut/ Updated: 2025-12-11 Registering the keyword as the domain worked in the era when term matching carried more weight, which is why the idea has such a long tail, and it worked regardless of whether the site was any good, which is why Google shipped an update in 2012 aimed at low-quality exact-match domains. Google has since described words in a URL as a very minor factor: any exact-match domain ranking well today is ranking on its content and links and would do about as well under another name. The belief persists partly because matching words appear emboldened in the result, a display behaviour applied to matching text anywhere in a snippet, which may help the click but did not decide the position. The costs are permanent: no brand to build on, a name that boxes the business into one service in one city, an association with the spam and lead-generation networks that use the pattern heavily, and harder press coverage. Keywords in the URL path are a different and milder thing: the gain is legibility, which has nothing to do with matching. ## Social signals are not a ranking factor URL: https://alinboatca.com/blog/social-signals-are-not-a-ranking-factor/ Updated: 2026-01-15 Google has said repeatedly it does not use social share counts as a ranking signal, and links from those platforms are nofollow or ugc almost without exception. Two practical problems block it working as sold: the platforms have restricted and repeatedly changed public access to engagement data, and engagement can be bought in bulk for very little, which is the property that makes any signal useless the moment it decides rankings. The correlation studies find is real and the wrong way round, since a piece good enough to be widely shared is usually good enough to be widely linked, and the links are the counted signal. What social does: distribution, which is how journalists and bloggers find work worth covering and is the mechanism behind most successful digital PR; growth in brand and name searches, which is much harder to fake; incidental discovery of new URLs; and consistent profiles feeding the sameAs entity data a language model uses to resolve who you are. Do not buy engagement. ## You can have more than one H1 URL: https://alinboatca.com/blog/you-can-have-more-than-one-h1/ Updated: 2026-01-22 The single-H1 rule comes from early HTML where the H1 was effectively the document title, and HTML5 sectioning made the strict reading stop matching how pages are built. Google has been explicit that multiple H1s are fine and that its systems work out page structure regardless, which also reveals that the markup is only a hint. The rule survives because counting H1 elements is trivially checkable and produces a confident red row, the same reason keyword density and word count persist. The heading faults that matter are invisible to that counter: headings chosen for how they look over what they mean, which destroys structure while the page still looks right; headings that describe document shape, say nothing about content and are therefore hard to lift; no headings at all, common on commercial pages; and a scrambled hierarchy. The real reason to get this right is accessibility, since screen reader users navigate by heading level, and the correct structure produces the SEO version for free. ## Crawl budget is probably not your problem URL: https://alinboatca.com/blog/crawl-budget-is-probably-not-your-problem/ Updated: 2026-01-29 Crawl budget is shorthand for how much a server can take and how much Google wants to crawl, and the second half is not a fixed allowance issued to you: it is largely an output of quality and structure. Google's guidance is that sites under a few thousand URLs are generally crawled efficiently and do not need to manage it; it becomes genuine above roughly a hundred thousand URLs or with fast-changing inventory. Server logs settle the question and nothing else does: filter to verified Googlebot over a few weeks and look for a large share of requests landing on parameter URLs, filter combinations, internal search results and paginated tails while money pages are fetched rarely. Genuine cases are faceted ecommerce, large publishers and directories, enterprise templates multiplying across hundreds of thousands of pages, and sites slow enough that Google throttles itself. The fix is reduction, since there is no setting to adjust and nobody to persuade. When a page is not indexed the Search Console status names the real conversation, usually internal linking or a content judgement. ## Publishing frequency is not a ranking factor URL: https://alinboatca.com/blog/publishing-frequency-is-not-a-ranking-factor/ Updated: 2026-02-05 Google does not reward a posting rate. The freshness behaviour people half-remember applies to queries where recency is part of relevance, such as news and prices. It does not apply to sites that publish often, so for an evergreen subject a thorough page from three years ago beats a thin one from Tuesday. The myth looks causal because teams publishing consistently are usually teams investing in content at all, and over a year some of those pages are good and earn links: the good pages did the work and the schedule was only the container. A cadence target has to be met whether or not anyone had something to say, which produces thin pages overlapping the ones already published, splitting authority through cannibalisation, lowering the site average and spending crawl attention. What outperforms it: updating pages already ranking at positions five to fifteen, which carry links and history a new page lacks; merging overlapping posts into one strong page and redirecting the rest; and publishing when there is something to say. Keep a rhythm as a working habit, drop it as a target. ## LSI keywords are not a thing URL: https://alinboatca.com/blog/lsi-keywords-are-not-a-thing/ Updated: 2026-05-14 Latent semantic indexing is a real technique patented in the late 1980s for small static document collections, and it does not scale to a live index of the web. Google has said publicly there is no such thing as LSI keywords and that it does not use LSI; the systems doing this job now are neural and work with meaning, entities and relationships, where LSI used a term-document matrix. The name stuck because a tool needed something authoritative-sounding for "related terms scraped from the pages currently ranking". The instinct underneath is sound, since pages covering a subject properly inevitably use its surrounding vocabulary, but that is evidence of understanding and no substitute for it: adding related terms does not make a page thorough any more than adding words makes it comprehensive. Treated as a checklist it produces sentences built to hold phrases and sub-headings promising coverage the page does not deliver. Use the lists as a gap check, asking whether an entry points at something missing, and otherwise write from actual understanding. ## A 301 redirect does not lose 15% of your link equity URL: https://alinboatca.com/blog/301-redirects-do-not-lose-link-equity/ Updated: 2026-05-28 The 15% figure is a misremembered version of the PageRank damping factor, which applied to links generally and was never a redirect tax, and Google has since confirmed that 30x redirects do not lose PageRank at all. Redirect type no longer affects equity but still decides meaning: a 301 says permanent so index the destination, a 302 says the original returns so keep it indexed, and a temporary redirect left for two years tells Google to keep a URL you abandoned. What costs you is elsewhere: redirecting to a non-equivalent page, which Google frequently treats as a soft 404 so the signal is discarded; chains, which add latency and collapse when a middle hop is removed; an incomplete map, which is the most common cause of a real post-migration drop; and leaving internal links, canonicals and sitemaps pointing at old URLs. Believing the 15% leads to worse decisions, such as refusing to consolidate cannibalising pages because merging feels like losing value. ## 404 errors do not hurt your rankings URL: https://alinboatca.com/blog/404-errors-do-not-hurt-your-rankings/ Updated: 2026-06-18 A 404 is the right response to a request for something that does not exist, Google expects every site of any age to return them, and the count carries no penalty. A site returning no 404s at all is more suspicious, because it usually means something answers 200 to everything. Four cases matter: a 404 on a URL carrying external links, the only one with a direct cost, which should be redirected to a real equivalent; internal links pointing at 404s, which are a bug, not the web being the web; a sudden spike, which almost always means a release dropped a path segment or a migration missed a rule; and soft 404s, pages returning 200 while saying nothing is here, which are the actively harmful version. Bulk-redirecting dead URLs to the homepage is worse than the 404, since Google reads a mass redirect to a non-equivalent page as a soft 404 anyway and the visitor lands somewhere that does not answer them. Use 410 where you have removed content permanently, and ignore the long tail. ## Domain age is not a ranking factor URL: https://alinboatca.com/blog/domain-age-is-not-a-ranking-factor/ Updated: 2026-07-02 Google has said the age of a domain confers nothing by itself, yet old domains rank better on average, and both are true because age stands in for what a domain collects while it sits there: links, useful content, name mentions, direct traffic and evidence of being a going concern. The cleanest disproof is a domain registered years ago and parked untouched, which is exactly as old and worth nothing. This is the sandbox myth from the other end, putting the cause in the calendar because a calendar asks nothing of you. Buying an expired domain usually disappoints because what transfers is a backlink profile: relevance rarely survives a change of subject, the pages those links point at are gone so they resolve to 404s or a homepage redirect that reads as a soft 404, prior spam history comes along too, and Google can treat a domain that changes ownership and purpose as effectively new, removing the only thing purchased. What compresses the timeline is earning the signals directly: a few earned links, pages that deserve indexing, internal structure that concentrates authority, and winnable target terms. ## Word count is not a ranking factor URL: https://alinboatca.com/blog/word-count-is-not-a-ranking-factor/ Updated: 2026-06-11 The 1,500 word minimum came from a correlation study, not from Google, which has said repeatedly there is no ideal article length. The correlation is real but points elsewhere: pages that cover a subject thoroughly tend to be longer, so length is a by-product sitting downstream where it is easy to measure. Reversing it into "make pages long" keeps the only part with no causal power, because you can add a thousand words in an afternoon but not understanding. Padding has predictable symptoms, a preamble and a history section and the answer at the seventieth percent, each a reason to leave before reaching it, and it works against extraction since both search engines and models reward a clear answer in a passage that stands alone. Thin content is a genuine problem, but the word count is only the symptom: the fix is giving the page something to say or merging it into the page that already does. ## An XML sitemap does not get your pages indexed URL: https://alinboatca.com/blog/xml-sitemaps-do-not-get-you-indexed/ Updated: 2026-06-25 A sitemap is a list of URLs with optional change hints, useful for discovery on large, poorly linked or brand new sites. Three decisions sit downstream of it and it has no say in any: whether to crawl, whether to index, and where to rank. Discovered currently not indexed proves the point, since it means the sitemap did its entire job and Google still declined to fetch the URL. Resubmitting repeats the same announcement and touches neither judgement, which is the sitemap version of pressing the lift button again, and bulk Request Indexing at volume is the same mistake. The one field carrying real information is lastmod, and only while it is true: stamping every URL with the build time, or with a future date, turns the field into machine noise and loses the per-site trust Google places in it. What gets a page indexed is internal links from pages already crawled, fewer competing URLs, an external link or two on a new domain, and the page being worth indexing at all. ## Google does not penalise AI content URL: https://alinboatca.com/blog/google-does-not-penalise-ai-content/ Updated: 2026-07-08 Google's stated position is that the guidelines are about content quality, however the content was produced. Automation making helpful original material is fine; generating volumes of pages that exist only to catch traffic is not, and that has been against the rules for over a decade as scaled content abuse, spun content and doorway pages. The feared classifier does not exist at the accuracy the fear requires: public detectors produce confident false positives on ordinary human writing and degrade with each model release, so a penalty built on them would misfire on legitimate publishers, and authorship method is a poor proxy when Google can assess the page directly. What goes wrong is ordinary and about substance: sameness, since a competent synthesis of the current top ten is by definition not additive and reads as crawled currently not indexed; confident errors; the absence of first-hand experience and specific numbers; and the volume the tools make trivial. Drafting, restructuring and summarising are real work made faster. The question is whether the page deserves to exist, not whether the tooling will be detected. ## Schema markup is not a ranking factor URL: https://alinboatca.com/blog/schema-markup-is-not-a-ranking-factor/ Updated: 2026-07-16 Google has said outright that structured data is not a ranking factor. It is a translation layer restating what the page already shows in a form a machine can parse, and it is used for two things: eligibility for rich results, where Google still decides per query whether to show the feature, and entity clarity so a machine can tell which organisation you are and who wrote what. The myth survives because adding markup is often followed by more clicks, which looks like a ranking gain in a report: the position did not move, the listing got better at being chosen from where it already sat. Four reasons valid markup earns nothing: a type with no rich result attached such as Organization, a type whose result was withdrawn such as FAQ since 2023, blocks left disconnected so nothing resolves into a graph, or markup contradicting the visible page, which is the actively harmful case and can reach a manual action. Review stars are where enforcement is real, since self-serving or fabricated ratings are exactly the pattern that gets checked. Worth doing for eligibility and legibility, and increasingly for how a model resolves who you are. ## Why we stopped adding FAQPage schema URL: https://alinboatca.com/blog/why-we-stopped-using-faqpage-schema/ Updated: 2026-07-14 Since Google restricted FAQ rich results to government and health sites in 2023, FAQPage markup earns most sites nothing. Why the code was removed instead of being maintained for a benefit that no longer exists, and when structured data is still worth adding. ## Technical SEO in 2026, in priority order URL: https://alinboatca.com/blog/technical-seo-guide-2026/ Updated: 2026-08-02 Every guide to technical SEO arrives as a list, which is the one shape the work cannot take, because a list with no order cannot tell you which three items are costing you traffic today. A missing canonical and an uncompressed image appear as equal line items when one removes a page from the index and the other costs a tenth of a second. The useful structure is four questions asked in sequence, each meaningless unless the previous one is answered yes. Can Google reach the page, which is about internal links more than sitemaps, status codes that match the content, and robots directives, and is almost never about crawl budget below a few thousand URLs. Is it indexed, where discovered and crawled but not indexed are different conversations and not degrees of one problem, and where noindex left after launch, canonicals pointing at redirects and parameter sprawl are the real faults. Can it be understood, covering rendering, where content that only exists after a client-side fetch makes an optional crawl step mandatory, and structure, where schema buys eligibility and entity clarity but not a position. And is there a reason to prefer it, which is where HTTPS, mobile and Core Web Vitals sit as genuine but small tiebreakers that will not lift a page failing question two. What changed for 2026 is narrower than the guides suggest: AI answers absorbing short factual queries, entity clarity mattering more because a model has no SERP to fall back on, and INP replacing FID. What did not change is word count, keyword density, publishing cadence and domain age, none of which were ever factors. The order is the method, and on a new domain the real constraint is often that nothing links to you yet. ## The structured data types that change your listing URL: https://alinboatca.com/blog/essential-structured-data-types/ Updated: 2026-07-28 Schema.org defines hundreds of types, Google renders a visible result for a small fraction, does quiet entity work with a few more, and ignores the rest. That gap matters because validating and earning something are two different tests and most guides only run the first. The types that earn a visible result are roughly seven: Product for price, availability and rating, BreadcrumbList which replaces the raw URL line and is the best ratio of benefit to effort on the list, Review and AggregateRating but only nested on a supported type and never on your own organisation, Event, Recipe, JobPosting and VideoObject. The types that produce no rich result and are still the most important markup on most sites are Organization, Person, LocalBusiness, Article and WebSite, because they answer who published this and what this business is, which is also what an AI assistant reads when deciding whether you are the same entity as a similarly named firm. The failure there is not absence but contradiction: three blocks declaring three slightly different organisations, or a LocalBusiness address that does not match the Business Profile. The dead types still shipping on many sites are FAQPage, restricted to government and health since 2023, HowTo, removed entirely, and the sitelinks search box, deprecated in 2024, all of which still validate perfectly. Markup contradicting the visible page is worse than no markup and is where manual actions happen. For most sites the whole job is three types: Organization or Person once with a stable @id, BreadcrumbList on every page in a hierarchy, and whichever single type matches what you sell. ## Mobile speed on WordPress is four problems URL: https://alinboatca.com/blog/mobile-speed-for-wordpress/ Updated: 2026-07-29 Advice on WordPress mobile speed is usually thirty tweaks, which is how sites end up with five optimisation plugins fighting each other and a slower site than before. In practice almost every slow WordPress site is slow for the same four reasons in a reliable order. Before any of it, check whether you are reading field data or the lab score: PageSpeed shows both, Google ranks on the field data from real Chrome users over 28 days, and the two can disagree sharply while both being correct. Mobile is slower because of network latency and a smaller budget for parsing and executing JavaScript, not because the phone is weak, so finding mobile slow means the page is heavier than you realised. Cause one is images, and it nearly always is: full-size uploads served to a 390px viewport, no modern format, no srcset, no dimensions so the layout jumps, and plugins that lazy-load the hero and thereby delay the exact paint being measured. Cause two is the theme and what it lets you build, testable by loading a nearly empty page and measuring it, because whatever that weighs is the floor every page pays. Cause three is plugin sprawl, including optimisation plugins stacked on each other, where the site is often faster with all of them off. Cause four is missing page caching, or caching that serves stale prices and hours invisibly. Be clear about the return: Core Web Vitals are a tiebreaker, so do this for the person waiting on a train with two bars of signal, which is a better reason and does not depend on an algorithm keeping its weighting. ## Best WordPress SEO plugins for 2026, compared URL: https://alinboatca.com/blog/wordpress-seo-plugins-compared/ Updated: 2026-07-31 Install Yoast, Rank Math, SEOPress or Slim SEO, change nothing, and view source: you get a title tag, meta description, canonical, Open Graph and Twitter tags, a robots directive, an XML sitemap and a block of JSON-LD. That is the entire job and all of them have done it competently for years, so the useful comparison is not about SEO capability but about weight, defaults, how much the interface nags you and what sits behind a paywall. Yoast is the heaviest with a good connected graph and redirects gated behind premium; Rank Math is moderate with the broadest free schema coverage; SEOPress is lighter with flat pricing that suits agencies; Slim SEO is lightest and decides almost everything for you, which costs nothing if it decided as you would have. The content analysis panel is the most misleading feature in the category: it scores against readability heuristics and a keyword you typed, none of it visible to Google, and writing to satisfy it produces the padded prose that density targets always produced. The setting that matters more than the plugin choice is what to do with author, date and tag archives, since the default determines whether a large number of thin overlapping URLs are indexable. The plugin is the problem in only three cases, all of them conflicts, none of them capability: two SEO plugins active at once producing two canonicals and two contradictory JSON-LD blocks, a theme emitting its own schema alongside the plugin, and the redirect manager, whose data is the hardest thing to take with you. If you are on the wrong one, stay on it: migration carries real risk and gains almost nothing. ## Five custom tools worth building for SEO and UX URL: https://alinboatca.com/blog/custom-tools-for-seo-and-ux/ Updated: 2026-08-01 Almost every list under this heading is a roundup of software to subscribe to, often with affiliate links. This is the other kind: tools built on your own site, a smaller category and a better investment when it works. A tool does two things a written page cannot, answering a query whose intent is to do something, and giving people a reason to link without being asked, which matters because on a site with no authority the constraint is never too few pages. Five patterns, cheapest first. The generator, a form and a template, ranks for queries no article can satisfy and is the most reliably linkable. The checker measures the thing people argue about and earns its place only by taking a position, which is why the meta description checker here measures pixel width instead of counting characters. The calculator turns "it depends" into a number, qualifying enquiries before they arrive, and is cheap to build and hard to agree internally. The finder turns data you already hold into an answer to a high-intent query and is where the commercial value is, illustrated by a near-me tool built for a client where the interesting engineering was a self-maintaining location layer so new locations need no code change, server-rendered because a tool a crawler cannot read earns nothing. The comparator removes the decision sending people elsewhere. Most custom tools fail the same way: built, published, nothing links to them, an expensive page with a maintenance cost. The two specific traps are orphaning, leaving the internal links until after publication, and building client-side only so the page ships empty HTML. ## How to improve Core Web Vitals in 2026 URL: https://alinboatca.com/blog/how-to-improve-core-web-vitals/ Updated: 2026-07-30 Three metrics with thresholds that must be met at the 75th percentile of real visits: LCP at 2.5s, INP at 200ms, CLS at 0.1. INP replaced FID in 2024 and was no rename but a genuine tightening, because FID measured only the delay before the browser began handling the first interaction while INP measures the full round trip to a visible response across every interaction, so sites that sailed through FID routinely fail INP. Use field data from real Chrome users over 28 days, since a page can score badly in the lab and pass in the field, in which case there is nothing to fix. LCP should be broken into four phases before touching anything, because the fix differs entirely by phase: time to first byte, which no image work will help; resource load delay, the phase almost nobody looks at, caused by the image being discovered late because it is set in CSS, injected by JavaScript or inside a carousel; resource load duration, the part everyone optimises; and element render delay, usually a blocking stylesheet or font. Lazy-loading the LCP image is the standard self-inflicted wound. INP is main-thread work, fixed by breaking up long tasks so the browser can respond between them, painting the visible part of a response before firing analytics, and auditing third-party scripts. CLS comes from three preventable things: images without dimensions, content injected above existing content, and fonts swapping with different metrics. Fix one thing, deploy, and wait, because field data moves on a 28-day window. Stop when the field data passes, not when the lab score is 100. ---