A UK trades job management and compliance platform wanted to know one thing: when a gas engineer or an electrician asks an AI engine which app to use, does the platform get named. Over four weeks, tracked against 27 platforms in the category including Tradify, Commusoft, Joblogic, Powered Now, simPRO, BigChange, ServiceM8, Fergus, Payaca and Jobber, its AI search visibility rose from 21.4% to 29.9%, roughly a 40% relative gain. That is the good number.
The other number belongs in this case study just as much. Average position, how early the platform is named within an answer once it is named at all, moved from 1.8 to 2.5 over the same four weeks. That is worse, not better. Both numbers are real, and a case study that only publishes the one that went up is not a case study. It is an advert.
TL;DR
- Visibility up roughly 40% relative in four weeks: from 21.4% (week of 22 June 2026) to 29.9% (week of 20 July 2026), through a dip to 19.6% along the way.
- Mentions grew from 1,285 to 2,265.
- Average position moved from 1.8 to 2.5 over the same period. That is worse, not better, and we are stating it plainly.
- The category itself changed shape during this exact window: AI search answers in this space carried sponsored placements for the first time.
| Metric | Value | Caption |
|---|---|---|
| Visibility | 21.4% to 29.9% | |
| Mentions | 1,285 to 2,265 | |
| Average position | 1.8 to 2.5, worse | |
| Platforms tracked in category | 27 |
What was the starting position?
In the week of 22 June 2026, the platform held 21.4% visibility against the 27 tracked platforms, was named 1,285 times, and averaged position 1.8 when it did appear, close to the top of the answer.
An audit of the platform's own site explained why growth was still available even from a reasonable starting point. There were zero comparison pages, zero alternative-to pages and zero category-level buying guides anywhere on the domain. The blog existed but never named a single competitor, which meant it never showed up for any of the comparison and alternative queries buyers actually type. There was no FAQ or product schema markup, no llms.txt file, and one product page was returning a server error instead of content. On top of that, an older brand domain the company had previously traded under was still live and had never been redirected, splitting years of accumulated authority across two separate web addresses instead of consolidating it on one.
Who else is answering these questions?
| Platform | Category |
|---|---|
| Tradify | All-in-one trade job management |
| Commusoft | Field service management |
| Joblogic | Job management and compliance |
| Powered Now | Job management, quoting and invoicing |
| simPRO | Enterprise trade job management |
| BigChange | Job management and CRM |
| ServiceM8 | Job management for small trades teams |
| Fergus | Job management for trades |
| Payaca | Quoting and job management |
| Jobber | Job management and scheduling |
Twenty-seven platforms in total, counting the one this case study is about. A long tail beyond the ten listed here fills out the rest of the category.
What did we actually do?
The audit pointed at a content vacuum, not a quality problem, so the response was a straightforward content programme plus a set of technical fixes.
Published new pages in four waves. The first wave, live the week of 13 July, targeted the clearest gaps the audit had flagged: head-to-head comparison pages against named competitors, "best app for" category guides for specific trades, and alternative-to pages for buyers already using a rival platform and searching for a switch. A second wave followed the week of 20 July, a third on 21 July, and a fourth on 27 July. The first two waves fall inside the measurement window covered by the numbers in this case study. The third and fourth continued the same work past it, so their full effect is not yet reflected here.
Fixed the schema and crawlability gaps. An llms.txt file went live, and Article and FAQPage structured data were added across the blog from 14 July, so pages that previously had no machine-readable structure at all now render cleanly for an AI crawler.
Flagged, but had not finished, the legacy domain fix. The unredirected old brand domain was identified as a priority-one issue in the audit. The redirect itself was still outstanding inside this measurement window. We are naming that here rather than implying it was already resolved.
Restyled the older content. Existing posts were brought in line with the new structural standard, rather than leaving a visibly two-tier blog where only the newest posts had comparison tables and FAQ blocks.
What actually moved, and what didn't?
| Metric | W/C 22 June 2026 | W/C 29 June 2026 | W/C 20 July 2026 |
|---|---|---|---|
| Visibility share | 21.4% | 19.6% | 29.9% |
| Mentions captured | 1,285 | not separately measured that week | 2,265 |
| Average position | 1.8 | not separately measured that week | 2.5 |
The dip in the second week is worth sitting with rather than skipping past. New content does not get crawled, retrieved and cited on the day it publishes, and the audit itself, plus the first wave of new pages, took most of that early window to land. Visibility fell before it rose.
| Category | Value |
|---|---|
| Week of 20 July 2026 | 2.5position |
| Week of 22 June 2026 | 1.8position |
Here is the part we are not going to soften. The platform is being named in far more answers than it was four weeks earlier, 2,265 mentions against 1,285. But when it is named, it now sits marginally further back in the answer than it did before, position 2.5 against position 1.8. Visibility and position are not the same metric, and they do not automatically move together.
Why was this exact window unusual for the category?
The content programme was not the only thing changing shape in this category during these four weeks. Sponsored placements began appearing in AI search answers for UK trades software for the first time during this exact period: zero ad-bearing answers across the first two tracked weeks, then a rapid acceleration by the end of the month. Tradify, Powered Now, Commusoft, Joblogic and Jobber were all bidding on competitor-alternative queries such as "best joblogic alternative uk" (we cover the ad activity itself, in more depth, in our separate reporting on this category's sponsored placements).
That timing matters for reading the position number honestly. Much of the new content published in this window was deliberately built for comparison and alternative-style queries, the same queries where a platform now gets named alongside several rivals in one answer rather than standing alone in a single-brand definitional answer. Being named in a crowded, multi-brand answer tends to land a platform further back than being the lone name in a narrow one. Some of the position movement is very plausibly a direct consequence of the exact strategy that drove the visibility gain: deliberately entering more contested, more competitive answers instead of only defending answers it already led.
What does a slightly worse average position actually mean here?
It means the platform traded some of its lead in narrow answers for a much larger footprint across contested ones. That is a real trade, not a rounding error, and it is worth naming as a trade rather than dressing it up as a second win. A platform that is invisible cannot lose position, because it was never there to begin with. A platform that shows up in more than 900 additional answers, in a category that just started running paid ads for the first time, and holds a position just past the leading two or three names, is in a materially stronger commercial spot than one that stayed narrowly ahead in a handful of quiet answers nobody was fighting over.
What does this mean for a trades platform thinking about AI search?
Three takeaways. First, a content vacuum is the easiest gap to close and usually the first place to look, comparison pages, alternative-to pages and category guides do the heaviest lifting when a domain has none. Second, expect a dip before a rise: new content needs to be crawled and retrieved before it can be cited, and judging a programme at the two-week mark rather than the six-week mark will make good work look like it failed. Third, watch both visibility and position, not just one. A platform chasing visibility alone, without checking what happens to position, could ship a strategy that grows the count of answers it appears in while quietly sliding backward inside each one.
FAQ
What is a UK trades job management and compliance platform, and why track its AI search visibility? These are software platforms tradespeople use to schedule jobs, invoice customers and issue compliance certificates such as gas safety records. Engineers and business owners increasingly ask AI engines which app to use rather than searching a traditional results page, so a platform's visibility, and where it gets named, now shapes real purchase decisions.
How much did this platform's AI search visibility change, and over what period? Visibility rose from 21.4% to 29.9% over four weeks, roughly a 40% relative gain, with a dip to 19.6% partway through. Mentions grew from 1,285 to 2,265.
Did every metric improve for this platform? No. Average position, how early the platform is named once it appears in an answer, moved from 1.8 to 2.5 over the same four weeks, which is worse. The platform is named far more often, but sits marginally later within the answers that name it.
What actually moved this platform's AI search visibility in this window? A four-wave content programme of comparison, alternative-to and category-guide pages targeting gaps a prior audit had identified, alongside adding llms.txt and FAQPage and Article schema across the blog from 14 July. A legacy, unredirected brand domain was flagged as a priority fix but was not yet resolved inside this measurement window.
Were AI search engines carrying paid ads in this category during this period? Yes, and for the first time. The category carried zero ad-bearing answers across the first two tracked weeks, then a rapid acceleration by the end of the window, with named competitors bidding directly on comparison and alternative queries.
If you want to see where your own category stands, the free AI search audit shows your current visibility across the AI engines your buyers actually ask.
