AI SEO & Marketing

Be found by search engines and by the AI that answers for them

Technical SEO and marketing engineering from the developers who also build retrieval systems, agents and knowledge bases. We understand what gets cited because we build the systems that cite.

Book a scoping call Start with an audit

Audits in 2–3 weeks
No ranking promises
Engineers, not account managers

How a page gets into an AI answer. Crawling and indexing are shared with classic search; AI engines then retrieve a handful of pages and cite them inside one synthesised answer.

What changed

Search now answers. Your page has to be worth quoting.

For twenty years the goal was a position in a list of links. Today a growing share of questions are answered directly by ChatGPT, Perplexity, Google AI Overviews and Claude-style assistants. Those answers are built by retrieving a few pages from an index and asking a model to synthesise them, with citations.

That retrieval step is the same mechanism we build for clients every week in AI knowledge bases and vector search. We know what a retriever rewards: clear entities, well-structured facts, fresh dates, clean HTML and markup that says what a page is about. We also know what it ignores.

So our SEO practice is not a content mill with an SEO plugin. It is senior developers applying the mechanics of retrieval, rendering and structured data to your site, and shipping the fixes themselves. This site runs on the same principles: build-time JSON-LD, a generated llms.txt, and pages written by people.

Signal	Classic search	AI answers
Unit of competition	A ranked position	Being retrieved and cited
What is rewarded	Relevance, links, speed	Entity clarity, structured facts, freshness
Winning format	Comprehensive articles	Tables, definitions, FAQs, original data
Success metric	Position, click-through	Citation share, mention share
Measured where	Search Console, rank trackers	Prompt-set monitoring you run yourself

Services

Fourteen services, three jobs

Be found by AI engines, keep the technical foundations sound, and turn the attention into inquiries. Most clients start with an audit and add services as the roadmap dictates.

Be found by AI

Generative Engine Optimization

Earn citations inside AI answers: entity clarity, quotable facts, structured markup and freshness.

Explore GEO

AI Search Visibility

Measure whether ChatGPT, Perplexity, AI Overviews and Claude mention or cite you, and track it monthly.

Measure visibility

llms.txt & AI Crawlers

A generated llms.txt, sensible robots.txt directives for AI bots and the trade-offs explained.

Set up llms.txt

AI Content Operations

Brief, research, draft with AI, expert edit, fact check, publish, refresh. Quality gates at every step.

See the workflow

Technical SEO

Technical SEO Audit

Crawl, index, render, Core Web Vitals, structured data, internal links and server logs, with a prioritised fix list.

Scope an audit

Structured Data

Schema.org markup generated from your CMS, validated, and linked into one graph machines can trust.

Implement schema

Core Web Vitals

LCP, INP and CLS diagnosed with field data and fixed in your framework by developers.

Fix performance

Programmatic SEO

Templates fed by real data, a unique-value test per page and quality gates before anything is indexed.

Scale responsibly

SEO for SaaS

ICP keyword maps, comparison and alternative pages, integration pages and docs that rank.

SaaS playbook

Local SEO

Google Business Profile, LocalBusiness schema and location pages for firms with several offices. We have five.

Local setup

Marketing engineering

Marketing Automation

Lead scoring, lifecycle journeys and CRM sync with Salesforce, built as systems rather than one-off campaigns.

Automate journeys

Conversion Optimisation

A hypothesis, test, learn loop; forms and CTAs that respect the visitor; tests with enough data to trust.

Improve conversion

AI Analytics

Attribution you can defend, dashboards that answer questions, anomaly alerts and privacy by design.

See the data

Email Marketing AI

Segmentation from behaviour, drafting with human review and deliverability basics done properly.

Email that lands

How an engagement runs

Audit first, then fix in sprints, then measure

We avoid retainers that pay for activity. Each cycle has a scoped list of changes, an owner and a measurement plan.

Audit
Crawl, server logs, Core Web Vitals field data, schema validation and a baseline of your presence in AI answers. Two to three weeks.
Prioritise
Every issue scored by severity and effort. You get a roadmap, not a 90-page PDF nobody reads.
Fix
Our developers ship the changes: templates, schema, performance, redirects, content structure. Reviewed in your repo.
Measure
Search Console, field performance data and a prompt-set monitor across AI engines. Reported monthly with plain commentary.
Iterate
Monthly cycles of content, markup and technical work, re-prioritised from the data. Stop whenever the backlog is empty.

Principles

How we keep this honest

Human-led, AI-assisted
AI drafts, researches and checks. People decide, edit and sign off. Nothing is published that a named expert has not read.
No ranking promises
We promise work and measurement. Positions and citations depend on systems we do not control and which change without notice.
We ship code
Recommendations become pull requests. Schema is generated, not pasted. Performance is fixed in the framework, not with a plugin that hides the problem.
Measure what matters
Inquiries, qualified leads and pipeline, with search and AI visibility as leading indicators. See AI analytics.

AI Knowledge Base (RAG)

The retrieval mechanism AI engines use, built for your own documents.

How retrieval works

GEO, defined

A plain-English glossary entry for generative engine optimisation.

Read the definition

Predictive Analytics

Lead scoring and forecasting models that feed marketing automation.

See the solution

Insights & Guides

Long-form, opinionated guides on RAG, evals, AI cost and governance.

Browse guides

FAQ

Frequently asked questions

Do you guarantee rankings or AI citations?

No, and you should be wary of anyone who does. Search engines and AI assistants change their systems without notice. We commit to specific, measurable work: issues fixed, pages shipped, schema validated, visibility tracked. We report results honestly, including the ones that did not move.

What is the difference between SEO and GEO?

Classic SEO earns a position in a ranked list of links. Generative engine optimisation (GEO) earns a citation inside a synthesised answer from ChatGPT, Perplexity, Google AI Overviews or Claude. The crawl and index are shared; the last step is different. See our GEO definition.

Do you write content with AI?

We use AI for research, outlines, first drafts and repurposing. A subject-matter expert edits every page and every factual claim is checked before publishing. We do not publish unedited generated text. Our content operations page explains the workflow.

How long does a technical SEO audit take?

Two to three weeks for a site of up to a few thousand pages, including log analysis and a prioritised fix list. Larger or multi-country sites take longer. We quote a fixed price after a short scoping call.

Can you fix the issues you find, or only report them?

We fix them. Our team ships code in React, PHP, Laravel, WordPress, Node.js and Python, so templates, structured data, performance work and redirects are done by the people who found the problem.