Clean technical foundations are the precondition for everything else. AEO sits on top of structured data. Content strategy sits on top of clean architecture. If your site has crawl issues, canonical chaos, or JavaScript that AI bots can't parse, nothing built on top of it will work properly. We fix the foundations. Twenty years running this for clients with serious scale.
Technical SEO is the discipline of making sure a site's architecture, code, and infrastructure let search engines and AI systems crawl, parse, and cite its content. In 2026 that means three things: clean HTML that AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) parse without running JavaScript; canonical and structured-data architecture that makes entity relationships explicit; and performance that doesn't penalise the site in either surface. It's the foundation AEO is built on, not a separate discipline.
Our overriding sense was that We Are All Connected were driven to make our collaboration work, to be the best it could be.
We run crawls with Screaming Frog and our own custom tools, benchmark against Ahrefs' technical indicators, and score the site against our technical SEO rubric: roughly 40 checks across crawlability, indexation, canonicals, JavaScript rendering, structured data, performance, internal linking, and AI-crawler accessibility. Every finding gets a severity score (critical, high, medium, low) and an implementation estimate.
We fix the crawl layer first, because nothing else matters if Googlebot or ClaudeBot can't reach the right pages. Robots.txt review, XML sitemap rebuild, canonical strategy, redirect-chain cleanup, pagination logic, and (for sites at scale) log-file analysis to see what's actually being crawled versus what we think is being crawled.
For sites built on React, Vue, Next.js, or any client-rendered stack, we audit what AI crawlers can actually see. Server-side rendering where possible. Static HTML fallbacks where not. Explicit robots.txt entries for AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) with a clear allow/disallow strategy. This is the single most-missed piece of modern technical SEO.
Organization schema for the company. Person schema for named authors. Article and BlogPosting for content. Product or Service schema where applicable. BreadcrumbList everywhere. FAQ schema where the content earns it. All wired together with consistent @id references so AI retrieval systems can parse the entity graph correctly.
We optimise for LCP under 2 seconds, CLS under 0.05, and HTML weight under 150kb for above-the-fold content. Image strategy, script loading, critical CSS, caching, and CDN configuration. At scale (PEI-size estates), this is where most of the real pipeline impact happens: a 2-second improvement in LCP across 500,000 pages is a different kind of material change than faster image loading on one landing page.
Internal linking sounds like a content-team job. At scale, it isn't. We audit internal-link equity flow, identify orphaned high-value pages, rebuild the topic-cluster link structure, and wire hub-and-spoke architecture into the site's physical information architecture. The result compounds across years, not months.
The classic technical-SEO issues (broken canonicals, crawl waste, duplicate content, thin pages) still matter. What changed is the addition of AI retrieval systems to the mix.
These aren't separate issues layered on top of traditional technical SEO. They're the reason technical SEO matters more in 2026 than it did three years ago.
40+ heuristic checks across crawlability, indexation, JavaScript rendering, structured data, performance, internal linking, and AI-crawler accessibility. Every finding tagged critical, high, medium, or low with an implementation estimate, so you triage by impact, not gut feel.
Robots.txt, XML sitemaps, and canonical strategy rebuilt from first principles. AI crawler access remediation across GPTBot, ClaudeBot, PerplexityBot, and Google-Extended. Server-side rendering for JavaScript-heavy sites, wired together with a documented structured-data estate and hub-and-spoke internal linking.
Performance optimisation with measured before-and-after baselines. Log-file analysis for sites above 10k pages. Monthly monitoring setup so regressions get caught before they hurt pipeline. Direct Slack access to the senior team, no account managers.
Senior technical SEOs who have run this for B2B estates at serious scale, not junior analysts working from a template. The same people who scope the audit do the implementation work and stay on the account.
PEI Group runs a large B2B financial media estate with more than 500,000 pages. We have supported the technical SEO and content architecture across crawl management, canonical strategy, structured data, and internal linking. When AI search started citing passages, the foundations were already in place: clear entity structure, clean canonicals, and parseable HTML.
A full technical audit of your site. Crawl analysis, indexation review, canonical and redirect mapping, JavaScript rendering audit, AI crawler access check, structured-data assessment, Core Web Vitals benchmarking, internal linking analysis, and a prioritised findings backlog. More than half of audits lead to an ongoing relationship.
The audit, plus implementation of the high-priority fixes. Crawl architecture, canonical rebuild, structured data, AI crawler access, and Core Web Vitals remediation. Full documentation and monitoring setup included.
Ongoing technical SEO work embedded with your team. Regression monitoring, new-template reviews, log-file analysis, migration support, and strategic input. Particularly valuable for sites at 50k+ pages.
Not sure which fits? Book a call.
Book a callMore relevant, not less. AEO depends on AI retrieval systems being able to crawl, parse, and extract your content cleanly. If your technical foundations are weak, AEO work on top produces inconsistent results. We audit and recommend technical SEO foundations first for most clients starting AEO work. The two disciplines are genuinely inseparable in 2026.
Yes, usually. The platform handles some technical basics automatically (sitemaps, canonicals, basic schema) but leaves important architecture decisions to the implementation. Internal linking, structured-data depth, performance optimisation, and AI crawler access are rarely handled well by platform defaults. At scale, platform-default technical SEO is almost always sub-optimal.
Case by case. For Next.js and similar frameworks, we audit what's actually being served to AI crawlers; many frameworks have subtle rendering modes that work for Googlebot and fail for ClaudeBot. Common remediation: server-side rendering for content pages, static HTML fallbacks for commercial pages, and explicit AI crawler allowances in robots.txt. Where the stack can't support SSR, we work with your development team to find the best compromise available.
Depends on the business. Most B2B companies we work with want to be cited by ChatGPT, Perplexity, and Google AI Mode, so we implement explicit allow rules for AI crawlers. For clients with sensitivity around training data (financial publishers, for example), we implement selective blocking: allow retrieval crawlers (Perplexity, ChatGPT search mode), block training crawlers (GPTBot, Google-Extended when used for training). It's a per-client strategic decision.
8-14 weeks, depending on site size and the depth of issues. Small sites with straightforward fixes can be 6-8 weeks. Enterprise sites with significant canonical or rendering issues are 12-16 weeks. We scope and commit to a fixed timeline after the audit, no surprise extensions.
Yes, and we recommend talking to us before you commit to a migration strategy, not after. We've worked on dozens of migrations: replatforms, domain consolidations, subdomain-to-subfolder moves, and headless rebuilds. Migrations done well maintain rankings and performance. Migrations done badly lose years of authority in months. The pre-migration architecture work is usually more valuable than the post-migration recovery work.
Yes, always. About 90% of our technical SEO implementation work involves a client development team doing the actual code changes, with our team scoping, specifying, and reviewing. We write implementation briefs, review PRs, test deployments, and ship the monitoring. Good technical SEO engagements are collaborative, not handed-over.
Tell us what's performing below expectations, what's breaking, and what you suspect is wrong. We'll tell you honestly whether a technical SEO engagement makes sense, or whether the fix is somewhere else.
Book a call