Being named is an observation
Mentions, citations and correct descriptions show whether buyers can find you in an answer. They say nothing about whether the right buyer then got in touch.
Industrial suppliers: more qualified RFQs →
Get found, evaluated and contacted by the right buyers.
Give us a commercial objective, a target buyer, a budget and permission to execute. We map the questions those buyers ask across search and AI answers, ship the work that fixes where you are absent or unconvincing, and measure the enquiries your sales team actually accepts.
Free demand snapshot · one buyer journey · about 12 questions · two surfaces · no call needed
Where we observe your buyers' questions, alongside Google and Bing search
- ChatGPT
- Claude
- Perplexity
- Gemini
- Google AI Overviews
The honest problem
A citation is not a lead.
It is easy to sell AI-visibility anxiety and hard to show it produced anything. We keep three kinds of evidence apart, because each one answers a different question.
A visit is not an enquiry
Relevant visits, quote starts, completed forms and calls show whether the journey works. Plenty of traffic is research that will never become a quote.
Only your team can accept a lead
Accepted opportunities, rejection reasons and wins are the numbers that justify a budget. We need your CRM feedback to report them, and we say so before we start.
Visibility still matters: a buyer who never sees you cannot contact you. It is a diagnostic underneath the numbers that justify a budget, not the result itself.
How a mandate works
You bring the commercial problem. We run the programme.
Not “please audit our website”. A short brief in your terms, with a budget and a definition of a good enquiry, that both sides can hold the work against.
Increase relevant quote requests for commercial generator rentals in our served regions. Focus on construction and facilities buyers. Work within a £6,000 monthly budget. Exclude domestic enquiries and unsupported locations. We can support additional contracts, respond to enquiries within one business day, and provide CRM feedback. Begin with a baseline and a bounded execution programme.
- 01
Agree the mandate
- 02
Map the buying questions
- 03
Find where you are absent or unconvincing
- 04
Assemble and approve the work
- 05
Execute with review gates
- 06
Measure accepted enquiries, then re-plan
Three stages
Free to qualify. Paid to decide. Budgeted to execute.
The snapshot shows whether there is a problem worth funding. The diagnostic tells you whether the investment makes sense. The programme does the work.
Demand Snapshot
One buyer journey, about 12 questions, two surfaces, observed more than once. It states its own limits and ends with the next commercial question, not a sales pitch.
FreeBack in a few working days
Request a snapshot →Demand Diagnostic
A bounded commercial and technical assessment of one buying journey: what your buyers ask, where you are absent or unconvincing, what a qualified enquiry is worth to you, and whether a programme is worth funding.
£750–£1,500One-off · about two weeks
Explore the Demand Diagnostic →Demand Programme
A budgeted monthly programme against an agreed commercial objective. We plan, assemble and ship the work, then measure the enquiries your sales team accepts, not just mentions.
Typically £4,000–£8,000 / monthMonthly · reviewed against accepted enquiries
Explore the Demand Programme →Prices are ex VAT and are ranges, not quotes. A programme is a fixed orchestration fee agreed in advance plus an execution allowance spent only on approved work; unused allowance is credited, returned or rolled forward. Large catalogues and fleets can add buying-question page builds as a work package. How pricing works.
Who it's for
Specialist businesses with a valuable offer and room for more of the right work
One direct segment we lead with, one we know well, and agencies who already serve them.
Where we start
Specialist industrial suppliers, rentals and distributors
A new depot or territory, underused fleet, a catalogue buyers cannot search, or enquiries that are mostly the wrong size. The mandate: more qualified RFQs for a defined product family and territory.
Industrial demand →AlsoTechnical software and API businesses
A launch, a new integration, or a competitor owning the evaluation. The mandate: win the relevant evaluation journeys and the opportunities sales accepts.
Technical B2B →Through agenciesAgencies with a client asking for this
Your client wants AI-search execution and your team is full. We run it for a named client, under your brand, with ownership written down.
Agency pilots →Good fit
- A valuable offer and an identifiable buyer
- Capacity to take on more of the right work
- Someone who can approve spend and implement changes
- Willing to tell us which enquiries were useful
Not a fit, and we will say so
- You want a guaranteed ChatGPT recommendation
- Nobody can describe what a good enquiry looks like
- Changes to the site cannot actually be made
- Lead quality will not be reported back
Our work
Demand systems we built and run
Buying-question sites in industrial and technical markets, each with a quote, sign-up or enquiry path on every page.
Oil & gas · Global
Oilfield drill string marketplace
Operators worldwide request quotes from a catalogue built to be the answer.
Generating inbound quote requests from operators and contractors.
Equipment rental · USEquipment rental rates platform
Rental rate questions answered with numbers, then turned into quotes.
Generating inbound rental quote requests.
Construction · US & CanadaConstruction matting marketplace
A niche nobody explained became a steady source of mat requests.
Generating inbound mat pricing requests, routed to suppliers.
Developer tools · GlobalEvidence API for AI agents
A technical product made legible to the buyers and agents that evaluate it.
Generating inbound developer sign-ups and enquiries.
Developer tools · GlobalNews API for AI agents
A crowded category, entered with specific capabilities and specific buyers.
Generating inbound API sign-ups and enquiries.
What we publish: what we built and how many pages, taken from each site's sitemap. What we do not: lead volumes, or any claim about what caused them. Names are withheld. All work.
How we report
Three layers, three labels, no blending
Every number in a monthly review says which layer it belongs to and how we know it. When the evidence cannot support a causal claim, the report says so.
What was directly recorded: the form, the referrer, the call log, the declared source.
What you or your buyer say helped. Useful, and labelled as testimony rather than measurement.
What a properly designed comparison supports. Often unavailable, and then we say so plainly.
“Accepted organic enquiries rose after these changes, but causality remains uncertain” is a more useful sentence than invented precision. See the methodology, and our own numbers, including the run that went down.
The platform
Agents assemble the work. People own the commitments.
The agents make a programme cheaper to run and faster to change. They do not sign contracts, publish unreviewed claims or spend money outside an approved mandate.
Demand researcher
Finds the accounts, buying questions and triggers worth pursuing.
Keeps facts separate from hypotheses.
Qualification agent
Turns enquiries and replies into structured mandate records.
Never invents budget or authority.
Planning agent
Proposes work against the commercial objective, not a content quota.
Respects budget, evidence and permissions.
Supply agent
Matches each approved work package to internal capacity or a vetted specialist.
Commits spend only inside an approved mandate.
Execution agents
Prepare pages, product data, schema and quote-flow changes.
Review gates, versioning and rollback.
Measurement agent
Reconciles observations, visits, enquiries and your accept or reject decisions.
Does not turn correlation into causation.
People own: commercial commitments, sensitive claims, publishing approvals, supplier exceptions and the relationship with you. Every action records who approved it, what it cost and which hypothesis it tested.
Explore the platformWorking in a particular market? We keep notes on 31 cities, including the named alternatives in each.
Frequently asked questions
What does Word of GPT actually do?
We take on a defined commercial problem (more qualified quote requests for a product and territory, say, or more relevant evaluations of an API) and run a budgeted programme to win the demand behind it. We map the questions your buyers ask in search and AI answers, find where you are absent or unconvincing, ship the pages, data and enquiry flows that fix it, and measure the enquiries your sales team accepts. SEO, AEO and GEO are the channels; the unit we work towards is a qualified enquiry.
Why not just improve our AI visibility?
Because a citation is not a lead. Being named in ChatGPT or Google AI Overviews is useful only if it puts you in front of a buyer who can become an accepted opportunity. We still measure mentions and citations, but as diagnostics underneath the numbers that matter: relevant visits, enquiries, and the enquiries your team accepts. A programme that raises visibility without producing relevant enquiries has a journey or offer problem, and we will say so.
What do you need from us to start?
A commercial objective, a target buyer, a budget and permission to execute. In practice that means someone who can approve spend, a clear description of a good and a bad enquiry, access to the website, analytics and CMS, and a willingness to tell us which leads were useful. If you cannot share CRM feedback, we can still run discovery work, but we cannot honestly report on outcomes.
Can you guarantee we will be recommended by ChatGPT or Google?
No, and nobody honestly can. Google states it does not guarantee crawling, indexing or serving even when a site follows its requirements, and no AI engine sells placement in its answers. What we are accountable for is the agreed execution, budget discipline and measurement. You see what we shipped, what it cost, and what happened afterwards.
What does it cost?
The demand snapshot is free and deliberately capped. The paid Demand Diagnostic is £750–£1,500. A Demand Programme typically runs £4,000–£8,000 a month: a fixed orchestration fee agreed in advance plus an execution allowance spent only on approved work, with unused allowance credited, returned or rolled forward. Prices are ex VAT and are ranges, not quotes.
Do you work alongside our existing SEO agency?
Yes. Often the agency owns the relationship and the site, and we run the buying-question research, execution and measurement for a named client under the agency's brand. Directly, we can take a defined slice of work and leave the rest where it is. Either way the scope, ownership and reporting are written down before we start.
Have a commercial problem worth funding?
Start with a free, capped snapshot of one buyer journey. If the economics hold, we scope a diagnostic or a programme against the enquiries you want more of.