Why these are not one page
Most agencies describe themselves as working with everyone and then prove it with a page that could be about anyone. It is a reasonable way to avoid turning work away and a poor way to win it, because the buyer reading it has a specific problem and can tell within a paragraph whether you have met it before.
The six verticals below are genuinely different pieces of work. For B2B SaaS, the hard part is economic: the queries worth the most have almost no search volume, the category may not have a name yet, and a three-month sales cycle makes last-click reporting erase the contribution of the page that started everything. The job is choosing the right low-volume targets and proving they paid.
For ecommerce, it is architectural. A catalogue with four filter dimensions generates more crawlable URLs than the store has products, variants split ranking signals, category pages cannibalise each other, and discontinued products take years of accumulated links down with them. One template decision applies to ten thousand pages, which is why the leverage sits in the structure rather than the copy.
For local and service-area businesses, the main asset is not the website at all. The Map Pack sits above the organic results and ranks on relevance, distance and prominence, driven by a Google Business Profile, reviews and citations. You can rank well organically and still get no calls, and your position changes from one postcode to the next, so there is no single number to report.
Three more sit on top of that foundation with constraints of their own. Healthcare is judged against expert consensus rather than against the competing page, which makes named clinical authorship the mechanism rather than a nicety. Law firms are usually carrying a practice-area and location matrix that is quietly competing with itself, so the first deliverable is a list of pages to delete. Home services are capped by travel radius before anything else, and most of the job is won from a profile before anyone opens the website.
What carries across all three is the craft: technical foundations, a content engine run by senior editors at AI speed, and the entity and structured-data work behind being cited by AI search. Those are documented under what we do, deliberately without an audience attached, because that is the part that genuinely does not change.
When a vertical earns its own page, and when it does not
Six is a deliberate number rather than a stopping point we reached by accident. Vertical pages are easy to mass-produce and the temptation is obvious: twenty of them looks like coverage, ranks for twenty head terms, and costs a fortnight. It is also how a site becomes a collection of near-identical pages that Google consolidates and a buyer sees through in one paragraph.
Our test is whether the vertical has a distinct structural problem we can describe specifically. Ecommerce has faceted navigation generating more URLs than the store has products. Local has a ranking factor nobody can optimise, because distance is fixed. Healthcare has content assessed against expert consensus rather than against the competing page. If we cannot name the problem without resorting to generalities, the vertical does not get a page, however attractive the search volume looks.
The second test is whether the advice would differ. A page that recommends the same sequence of work as the page next to it is not a vertical page; it is the capability page with a different noun at the top. We check this mechanically before publishing, comparing the rendered sections across all six, and if a section is identical on two pages it gets rewritten or removed.
There are verticals we have deliberately not built. Financial services, education, manufacturing and hospitality all have genuine search dynamics, and we could write plausible pages about each. We have not, because plausible is the problem: without real engagements behind them, those pages would be competent summaries of other people’s writing, which is precisely the thing this site exists not to produce.
What does not change between them
The six playbooks differ in sequence and emphasis, not in craft. Every one of them rests on the same foundation: pages that can be crawled and indexed, a site that renders content server-side so AI crawlers can read it, an entity graph that resolves consistently, internal links that concentrate authority where it converts, and content with something true and specific in it.
That shared layer is documented separately, without an audience attached, under what we do. It is written that way on purpose. The technical work on a clinic and on a catalogue is substantially the same work; what differs is which part of it is the constraint, and in what order it pays to fix things.
Where our proof comes from
One engagement is measured, cross-checkable and ours to cite: sendr.ai, a B2B SaaS company taken from zero to 1.05M organic impressions and 7,430 clicks in six months, and to the #2 position in Google's AI Overview for its category query, above ZoomInfo. You can read the full case study.
That result is in one vertical, so it is cited on one page. The ecommerce and local pages carry no results section, because we do not yet have results in those verticals and borrowing a number from an adjacent one would be the first thing we tell clients never to do. Those pages argue from method instead. If a vertical-specific reference is a requirement for you, that is a fair requirement, and today we would not meet it.