Platform
Fortitude SentinelTrack how four AI engines recommend you and your rivals Fortitude ForgeThe team that moves your score, content, PR and web The VaultYour category's history, recorded before you arrived The whole platformHow Sentinel, Forge and the Vault fit together→
Why AI
Pricing
Insights

Pillars
AI & LLM OptimisationHow AI engines choose who to recommend13 Content StrategyTopic clusters and content AI cites12 Online PR & AuthorityEarned signals AI engines trust11 Website & PerformanceArchitecture for AI crawlers and buyers14 Industry GuidesVertical playbooks by sector16 From the VaultLive data reports on who AI recommendsLIVE The libraryAll 92 insightsSeven pillars plus weekly Vault reports.
About

The receipts for the AI age.

A continuous, week-by-week record of how ChatGPT, Claude, Gemini and Google AI Overviews answer the questions your buyers actually ask, across the top B2B brands in the UK, Europe and the United States. Your category's past, kept properly, so the day you arrive is not day one.

See a sample Vault report → Run the free check

What is in the Vault

The only AI visibility dataset that started before you did.

The Vault is not a logging tool that switches on when you sign up. It is a research index, run by us on a fixed cadence, across the brands and categories that actually matter to B2B buyers.

Four engines, every week

Every prompt is run against ChatGPT, Claude, Gemini and Google AI Overviews on the same day, so engine drift is visible rather than buried in averages. We keep the raw answer, the citations and the timestamp.

Thousands of brands, named or not

We index the top B2B brands in our coverage categories whether or not they are paying us. Your past is in the Vault before you arrive, and so is your rivals'.

Buyer questions, not vanity prompts

A curated library of the questions real buyers ask at each stage of the decision, refreshed quarterly with evidence from sales calls and search demand. No "is brand X the best CRM" theatre.

Scope

Indexing the top B2B brands across the UK, Europe and the USA.

Coverage rolls out market by market, category by category. If you do not see your category, ask. We add categories on demand and most can be backfilled inside a quarter.

Live now

United Kingdom

B2B SaaS, professional services, manufacturing, fintech, legal, healthtech, recruitment, logistics, agencies. The brands shortlists are built from, named or not.

Rolling out

Europe

Ireland and the Nordics live, with DACH, France, Benelux and Iberia added through the year. Prompts adapted to local buying language and procurement norms, not translated word for word.

Rolling out

United States

East Coast metros first, then West Coast and central. Same engine line-up, same cadence, same answer-shaped questions, with categories prioritised by client demand.

Coverage, brand counts and category lists are refreshed monthly. Public reports state the exact scope on the day they were drawn.

Honest by design

The model is Ahrefs, not surveillance.

Your prompts, your competitors, your settings stay yours. The Vault only records answers to our own canonical question library, run by us, on a fixed cadence. That is what lets us publish category reports without selling out the people inside them, and what lets you trust the chart when your name finally appears.

Open methodology, published
Same engines, same questions, same week
No client data sold or shared
Re-runs on request when an engine ships a major update

What we will never do with the Vault

The shortlist of things that make this dataset usable rather than another data broker.

Sell prompt-level data to your rivalsnever

Mix client prompts into the public indexnever

Hide methodology behind an NDAnever

Quietly change the question set between runsnever

Backfill answers we did not actually observenever

What it is for

Three ways teams use the Vault on day one.

Instant backfill

Up to twelve months of category history on your dashboard the moment you sign up. The conversation starts at, here is what has been happening, not, come back in a quarter.

Competitor and target diligence

Watch a competitor's AI footprint, or a target acquisition's, over time and across four engines. Useful before an offer, a pitch, or a board paper.

Market reports that write themselves

Quarterly category indexes and monthly movers reports are drafted from the Vault automatically and edited by humans. Some are public. Some only Forge clients see.

Methodology, in plain English

How the Vault is actually built.

The question library. Each category has a curated set of buyer questions, written from real sales calls, search demand and procurement RFPs. We do not test "what is the best X". We test the questions buyers actually type.

The engine run. Every prompt goes to ChatGPT, Claude, Gemini and Google AI Overviews in the same window, weekly, from a clean session, with the same parameters. We keep the raw answer, the citations and the timestamp.

Brand resolution. We resolve mentions to entities, not strings, so "Acme", "Acme Ltd" and "acme.com" land in one place. We tag whether the mention was named, listed, recommended, or hedged.

Share of voice. Computed per engine, per question stage, per market. We publish the formula and the edge cases, because anyone selling you a single "AI visibility score" without showing how is selling you a chart, not a dataset.

Refresh and re-runs. Categories are reviewed quarterly. When an engine ships a major model update we re-run the affected weeks on request, so a comparison across the change is honest.

Read the full methodology →

Vault reports

Some of the Vault is public.

Every month and every quarter we draft category reports straight from the Vault. The public ones live in Insights. The deeper ones go to Forge clients.

Quarterly category indexes

Who AI engines named, in what order, in your category, in the last 90 days. With sources, prompts and movement against last quarter.

Monthly movers

The brands that climbed, the brands that fell, and what changed in their published footprint that explains it. Drafted from the Vault, edited by humans.

Engine drift notes

When ChatGPT, Claude, Gemini or Google AI Overviews changes its mind, we say so, with the before and after answer side by side.

Read the latest reports → Browse by industry

Your category, already on your chart

Start with the free check. See your Vault history on the next screen.

Two minutes to your AI visibility score in four engines, with the last twelve months of category history loaded the moment you sign up.

Run the free check → See pricing Talk to us about a category