Monthly service
Hand us the work, every month.
A report tells you what is wrong. The monthly service is us putting it right: every month we measure again, find what moved and why, do the work on your site and beyond it, and prove it with the next measurement. From £149 a month, rolling, cancel any time.
Not sure it is for you? The free audit shows where you stand in a few minutes — and if the engines already name you everywhere, we will tell you so and you should not buy this.
Why you are not in the answer
There are five reasons. Yours is one of them, and it is knowable.
This is not a checklist we apply to everyone. The system decides which of these is true for each individual question you are missing from, and shows the measurement it used to decide.
The engines cannot read you
Your robots.txt blocks the AI crawlers, your sitemap is broken, or the pages that answer the question render only in the browser. Nothing else matters until this is fixed, so it is the first thing we check and the first thing we fix.
crawler_blocked · sitemap_unhealthy
No page on your site answers the question
Buyers ask about a named procedure, a price, a location, a comparison. If there is no page for it, the engine takes the answer from someone who has one. We find which questions have no page behind them.
no_page_for_topic · page_too_thin
The answer is there, but not in a form they can lift
The fact is in paragraph nine, or in an image, or split across a PDF. Engines quote what they can extract in one pass: a direct opening answer, a list, a table, a marked-up FAQ.
no_answer_format · missing_schema
They never read your site at all
The answer was assembled from directories, review sites and press — and you are not in the sources they used. We show which URLs they actually cited, then work on the ones you are missing from.
not_cited · third_party_absent
What others say about you contradicts your own site
An old price in a directory, hours that changed, a service you no longer offer. The engine resolves the conflict by picking one, and it is not always yours. Wrong pricing is the most expensive version of this.
narrative_contradiction
What happens every month
Measure, diagnose, plan, do, verify.
The fifth step is the one most of this industry leaves out. Work that cannot be checked against a measurement is an invoice, not a result.
Measure
Fifty purchase-intent questions, four engines, three runs each — 600 answers. Three runs is the point: ask an engine the same thing twice and you can get two different lists, so a single run tells you whether you appear, not how reliably. Every answer is stored raw with the URLs it cited.
Diagnose
Each question you are missing from gets a cause, not an opinion — and the evidence behind it: the crawl of your page, the sources the engine used, what your competitors have that you do not. Questions where you are already named are marked defended and left alone.
Plan
The causes become a ranked list of jobs, each with an expected point gain and an hours estimate, split into sprints. Ranking is by what it is worth against what it costs, so the top of the list is the cheapest real gain, not the biggest idea.
Do the work
On Monitor, you or your team get the brief and do it. On Autopilot, we apply it to your site through the access you grant: schema, llms.txt, robots, sitemap, internal links, answer formatting, new pages written to the brief. Every change is logged and reversible.
Verify
The next run checks the work. Each job carries the measurement that should move if it worked, and the report says plainly whether it did — worked, no change yet, or backfired. Work that does not move the number stops being recommended.
What lands on your desk
Six things, every month, in your account.
Your dashboard
Score and movement, the questions you won and lost this month, the cause behind each gap, and the plan. Updated with every run, not emailed once a month.
A monthly report you can forward
One page of what moved and why, then the detail. Written so a partner or a board can read it without you translating it first.
The prompt library
Every question with its funnel stage, visibility rate and variance band across the three runs. This is the raw material — you can audit our conclusions against it.
The change log
Every change made to your site, when, by whom, with the before and after, and a one-click reversal. On Monitor this is the brief you were handed instead.
Citation and source map
The real URLs the engines read for your category, which of them mention your competitors, and which of them you are missing from.
Third-party corrections
Wrong prices and stale details in the directories the engines actually read, drafted and tracked to resolution rather than listed as a recommendation.
Proof
We publish measurements of the best-known names in each sector.
The usual proof in this category is an agency measuring itself and anonymising everyone else. So instead we measure whole sectors in public, with names and dates, and you can run the same measurement on your own domain for free and check that it behaves the way we say it does.
Plans
The difference is who does the work.
The measurement is identical on all three — same fifty questions, same four engines, same three runs. What changes is whether the plan lands on your desk or gets done for you.
Card required. Cancel before day 7 and you pay nothing.
Then billed monthly. Cancel any time. Your measurements and dashboard stay yours after you stop.
Before you buy
What we will not promise you.
- Not a guaranteed mention. We do not control the engines. Anyone who guarantees a place in an AI answer is either guessing or gaming something that will be patched.
- Not what you see in the ChatGPT app. We measure through the engines' APIs with web search on. That is close to the consumer apps but not identical — personalisation and memory differ. Every report says so.
- Not a fast number. The technical work shows up within a run or two. Being added to the sources engines read takes months, and it is usually where the real gain is.
- Not bought links or fake reviews. Corrections go to the source, in your name, with the true facts. It is slower and it survives.
- Not a lock-in. Monthly rolling, cancel from your dashboard. Everything measured stays yours and stays exportable.
Anything here you want argued properly rather than in a bullet? Write to hello@namedfirst.co.uk.