What actually happens in the first 90 days — and what decides it.
A week-by-week account of how a cohort build behaves once it is live: what moves first, what compounds, what takes patience, and the three things that separate a firm that gets results from one that gets a nice website.
Why this matters: any agency can write "increased enquiries by 240%" on a website. Very few can tell you what will move in week two, what will not move until week six, and what will quietly decide whether the whole thing works. That is what this page is for.
First, the honest expectations
- Nothing happens in the first fortnight except the build. That is the work — 300+ pages do not write themselves.
- Visibility moves before enquiries do. Pages start ranking in weeks 3–6; the phone follows, because a ranking page on a Tuesday gets its first call on a Wednesday.
- Intake speed changes conversion immediately — the day the AI layer goes live, the same enquiries convert better. It is the fastest-acting piece.
- AI citations and reviews compound quietly. They do very little in month one and a great deal by month six.
- Your behaviour decides the ceiling. Firms that answer fast and follow up properly get materially more out of identical builds. That is not a disclaimer; it is the single biggest variable.
The build: nothing visible, everything decided
The first fortnight is strategy and production. Competitor scan, practice-area towns, sitemap, copy and the design direction — which you approve before anything is built. By day 14 the site is live with several hundred pages published, tracked and populated.
Search engines start crawling almost immediately. This is the least rewarding and most important phase: the ceiling for everything afterwards is set by which combinations you chose to cover.
Most common mistake here: narrowing the town list to "the places we already get work from". The build's advantage is coverage. Firms that trim it to three towns cap their own result before launch.
The invisibility starts to lift
Specific pages begin to appear — first in the long tail, then for the combinations with real intent. You will see impressions and clicks climbing in Search Console before your fee-earners notice a change in the phone.
Simultaneously the intake layer is live: chatbot answering at 9pm, phone answered at 9pm, enquiries qualified and summarised before anyone reads them. This is where the first measurable difference usually shows — not more enquiries yet, but a higher share of the enquiries you already had turning into conversations.
It compounds — and the phone starts telling you
Rankings for higher-intent searches begin to hold. Reviews have started accumulating from the automation. AI assistants start citing the properly-structured pages when people ask for a solicitor in your area — the least-visible but most durable gain, because it is not something you can buy your way into.
At the end of 90 days the picture is usually: materially more search visibility across your towns, every enquiry answered within a minute at any hour, qualified before it reaches a fee-earner, and a dashboard that tells you which of it produced work.
The verified numbers go here.
Two slots, reserved for the first cohort firms. Each will carry real, consented figures at the 90-day mark — the same numbers from the monthly report they receive.
Reserved. Not filled with invented numbers.
Reserved. Not filled with invented numbers.
What I will publish, and what I will not. I will publish the monthly report figures — enquiries, calls, chats, attributed matters, and where they came from. I will not publish percentage leaps with no baseline, "up to 3x" claims, or screenshots of graphs with the axis removed. If a build underperforms, that goes on the page too, with the reason. Firms buy the second build from people who were straight with them about the first.
Three things decide the outcome
The build that covers 300 combinations outperforms the beautifully designed site that covers five, every time. It is unglamorous, it is the mechanism, and it is why "we will write content later" quietly becomes never.
Not your fees, not your website's colour. A five-minute reply converts dramatically better than a nine-hour one — and the AI layer makes the first response instant even when nobody is awake. Firms that leave this to office hours are handing work to the firms that do not.
The moment a firm can see that 40 enquiries arrived and 11 became matters, the conversation stops being about marketing and starts being about conversion. That shift — from guessing to knowing — is worth more than any single feature in the build.
The sceptic's questions
"How do you know this trajectory will hold for my firm?"
"My city is competitive. Does that change things?"
"How much of this depends on us?"
"Why publish placeholders instead of testimonials?"
"What if the numbers are not there at 90 days?"
You can be one of the two firms whose numbers end up on this page.
That is the trade the cohort price is built on: a £4,945 build for £400, in exchange for being willing to be the case study — with your real figures, published honestly, good or bad.
P.S. — If you would rather see it in action before spending anything, take the free audit. Two pages, your real numbers, one business day. That is the sensible first step, and I would rather you take it than buy on enthusiasm.
P.P.S. — When the first cohort firms' figures are published here, the £400 price will not exist for their towns and practice areas. That is not a threat, it is the arithmetic of one firm per specialism per city.
Start with the diagnosis
Two pages, researched by hand: where you rank today, what AI assistants say about your firm, the three competitors beating you, and the trajectory I would expect for your practice area and towns — with the assumptions written down.