01 · Probe
Ask the engines.
600+ buyer-grade prompts hit ChatGPT, Claude, Gemini, Meta AI, DeepSeek and Grok — in your language, your market, your category.
Technology
The pipeline that puts your name in the answer.
The pipeline
Four steps from invisible to named.
One pipeline runs the whole job — probe what AI says today, score the gap, ship the optimisation, measure the lift. Automation keeps it fast; team review keeps it grounded.
02 · Score
Read the gap.
Every answer is scored on visibility, accuracy, structure and authority. One number per engine, one combined index for your brand.
03 · Optimise
Optimise the business.
A ranked roadmap of content, structure, citation and on-site moves — generated from your data and reviewed before delivery, so the engines start naming you next.
04 · Re-scan
Measure the climb.
Thirty days after the optimisation ships, we run the exact same probe — and show, prompt by prompt, where the engines now place your name in the answer.
Inside the probe
600 prompts. Every engine.
The probe is how we learn what AI says about you today. We don't ask the engines what we want to hear — we ask them what your buyers type. The prompts are refreshed every cycle so the gap we work on reflects the questions that influence whether your name gets named.
- Prompts mirror intent. Comparison, recommendation, troubleshooting, pricing — the questions a real buyer asks before they choose a brand.
- Diversity is designed in. Long-tail, branded, unbranded, locale-specific — so one weak slot doesn't fake a strong score.
- Fresh every probe. Engines update their answers often. We refresh the prompt pack on every run so the read stays current.
Inside the score
One index, four honest signals.
The score is the map of why AI doesn't name you yet. Every answer the engines return is graded on four components — each scored 0–100, each tied to a specific lever we move in the optimisation.
Visibility
76
Do they name you?
How often the engine surfaces your brand on the prompts a real buyer would type — across the full prompt pack, not the easy ones.
Accuracy
82
Do they describe you right?
Whether the engine's description matches your positioning, audience and features — or quietly drifts into a competitor's pitch.
Structure
64
Can they parse you?
Schema, headings, entity markup and content extraction — the plumbing engines actually read when they decide to quote you.
Authority
71
Do they trust you?
Third-party citations, mention diversity and link signals — the proof engines look for before they put your name in the answer.
Inside the optimisation
Four levers. Every move ranked.
The optimisation is the product. Each item is a concrete move with a target, a reason and an expected lift — sorted so the change that pulls your name into the answer fastest ships first.
Content moves
Rewrite the paragraphs engines lift from. Sharpen positioning sentences, anchor your name to the buyer phrases that win the answer.
Pages · Sentences · PhrasingStructured data
Schema, entity markup, FAQ blocks and product graphs — wired so engines read your site the way you read it, not the way they guess.
Schema · Entities · MarkupCitation moves
Earn the third-party mentions engines treat as proof — directories, category lists, comparison pages and the publications your buyers already read.
Mentions · Listings · PressOn-site moves
Speed, crawlability, internal links and canonical signals — the quiet plumbing that decides whether engines bother reading the pages that earn you the answer.
Speed · Crawl · Canonical
Inside the re-scan
30 days later. Same probe. New answer.
Engines need time to read the optimisation, re-crawl citations and re-rank answers. Thirty days is a practical window for the signal to settle — long enough to measure, short enough to act on.
- Identical prompt pack. Same 600+ questions. Same locales. The delta is honest because the test is the same.
- Engine-by-engine breakdown. See exactly where you climbed — ChatGPT, Claude, Gemini, Meta AI, DeepSeek, Grok.
- The answers, side by side. Before and after on every prompt that matters, so the lift isn't just a chart — it's a paragraph.
Built for change
AI engines move. Your signals hold.
Models ship new weights. Answer formats change. New assistants launch. The optimisation we ship is model-agnostic — anchored in the structures and citations every major engine learns from, not the quirks of any single one. We re-tune as the engines move, so today's gains have a better chance of holding tomorrow.
Model-agnostic signals
Re-tuned on every release
Six engines, one playbook
Buyers are asking AI. Be the answer.
Six engines. One probe. An AI Visibility Sprint that turns your existing business, services and proof into answer-ready context, then a 30-day re-scan to measure the lift.
Start with your business £34.99 · AI Visibility Sprint · 30-day re-scan included