The Pipeline Engine · the monthly programme
Four workstreams, one report a month. The same buyer questions re-run on a fixed date and screenshotted, next to the pages published, the listings gone live and the enquiries that arrived.
The Pipeline Engine is a monthly programme that gets a UK power and energy engineering firm shortlisted when its buyers ask AI assistants and search who to use, and reports the result in dated screenshots and enquiries. It runs from £3,000 a month, three months then rolling, with the rate locked for twelve months for founding clients. It begins with the Pipeline Diagnostic, whose fee is credited in full against month one.
92 per cent of B2B buyers say AI shaped their supplier shortlist. Across manufacturing queries, manufacturers hold four of the fifteen most-cited domains; directories, distributors and publishers hold the rest. A firm that only describes itself is the first thing an assistant sets aside when a third party appears. The Engine does the work that changes which names come back, and measures it the same way every month.
Sources: Semrush, 622 B2B buyers, July 2026; Semrush data reported by PPC Land, September 2026.
What the programme does
Built from what your engineers already know. Each cell says what lands on your desk.
01
The five buyer questions from your Diagnostic, run on a fixed date each month through ChatGPT, Perplexity, Microsoft Copilot and Google AI. Every answer screenshotted and recorded as named, cited or absent against the baseline.
What you receive: the monthly report, with every screenshot, the sources behind each answer, and the movement since month one.
02
Pages that answer the question a buyer actually types: which obsolete drive, which fault, which connection constraint. Built from two hours a month with one of your engineers and checked by them, so nothing overpromises.
What you receive: published pages on your own domain, each with the question it exists to answer and the structured data underneath.
03
Your firm's name, location and capabilities consistent everywhere an assistant looks, and structured so a machine can read them: the site, LinkedIn, Companies House, the registers and every directory that lists you.
What you receive: validated structured data, one canonical description of the firm, and a crawler access check with each fix logged.
04
A claim on your own site is marketing. The same claim on a trade title, an association listing, an accreditation register or a published case study is evidence, and it is what an assistant reaches for first.
What you receive: live listings and mentions, each linked in the monthly report, in the places your buyers and their assistants already look.
The monthly evidence
The report has two halves: what the assistants said this month, and what arrived in your inbox. Neither is an impression count.
The price is written here, not quoted on a call. The founding rate is for the first three clients and carries founding terms in both directions.
One won contract against the monthly fee is the only arithmetic that matters, and the Diagnostic works it out from your own quote history before you commit to a month. Nobody is sold the Engine cold: it starts with the Diagnostic, or with the free snapshot before that. How every number in the report is measured is on how we measure, and the programme's method is described in full under AI search visibility.
Straight answers
From £3,000 a month, on a three-month minimum term and rolling monthly after that. Founding clients have the rate locked for twelve months. From client four the rate is £3,500 to £4,500 a month on a six-month term, set by the width of the buyer-question map. The Pipeline Diagnostic fee is credited in full against month one.
Answer pages take weeks to be crawled and months to be cited, and a directory or trade-press listing takes time to go live. Three months is the shortest period in which the monthly measure can show movement that is more than noise. After the third month it rolls month to month, and you leave with every page, listing and report.
The same buyer questions re-run on a fixed date across ChatGPT, Perplexity, Microsoft Copilot and Google AI, every answer screenshotted and recorded as named, cited or absent against your baseline; the pages published, the data fixed and the listings gone live that month; and the enquiries that arrived, with where each came from wherever that can be known.
No, and anyone who does is guessing. Assistants change their sources and answers without notice. What is guaranteed is the method: fixed questions, fixed dates, screenshots, and a baseline you saw before you paid. If the measure has not moved by the end of the third month, the review says so and you can stop.
About two hours a month with one engineer who does the work, so that the answer pages say what your firm actually does and nothing more. The measuring, the schema and entity work, the directory, register and press submissions and the monthly report are all done by us.
Related: the Pipeline Diagnostic it starts with · how we measure · who we work with
Where to start
Ten working days, £2,500 at the founding rate, credited in full against month one. Or start with the free snapshot and see who is shortlisted instead of you.