All resources
Delia Rowland
Case study
ServiceTitan is one of the best-known names in software for trade-based businesses. But AI visibility only counts when it's measuring the questions that matter to your buyers; and on the enterprise commercial-contractor questions, ServiceTitan wasn't consistently showing up. And where it did appear LLMs were placing them mid-answer rather than towards the top behind rivals like BuildOps.
So, that was the gap. We split the work in two: ServiceTitan kept building what it could control on its own site, and Noble took the part it couldn't: getting named in the third-party articles that LLMs cite. This is critical when you consider that offsite sources account for over 85% of what AI uses to shape answers.
The first move was finding the right questions. Noble tracked the specific non-branded prompts ServiceTitan's enterprise buyers were putting to AI, then identified the independent sources those answers were built from. Non-branded means the buyer names no company, so the AI builds its shortlist from whatever independent sources it trusts.
Over roughly two months, Noble placed ServiceTitan in 17 third-party articles. Visibility was tracked in Scrunch, which records how AI assistants answer a defined set of buying questions and how often (and where) a brand is named.
Results at a glance (April 3 – June 19, 2026: first ~12 weeks, ongoing engagement)
Position: +50% (relative) - where the AI positioned ServiceTitan in the answer as a 4th option, it now sits around 2nd on average
Citation rate: +36% (relative) - how often an AI tool links back to ServiceTitan’s website when answering questions
Enterprise 50+ technician cluster: citations +250%, presence +30% (relative) - the highest-value question, moving the most
The clearest movement is in position. Meaning not just whether ServiceTitan appears, but where in the answer it lands. Over the measured window, position improved 50% (relative): on a question where the AI ranked it around fourth before, it now ranks it around second. It holds the top slot in 65% of the answers where it appears, and the bottom slot in just 1%.
Citation rate moved alongside it, up 36% (relative). This is where a specific kind of placement earns its keep: linked brand mentions, where the third-party article naming ServiceTitan also points back to it. Those are the placements that most reliably register as a citation signal, and they're a core part of what Noble places.
Zoom into the hardest, highest-value question in the set: enterprise commercial contractors with 50-plus technicians and the movement is significant. On that cluster, across 194 responses:
The timing is consistent with a clear hypothesis: the run of commercial-contractor placements clustered in June, and presence on this question firmed to a steady 75%-plus in the weeks that followed, roughly the two-to-three-week gap these sources typically take to index. While we can’t know exactly when the model indexes the content, this hypothesis is consistent with research we conducted on brand mentions and the time to lift.
In the few weeks after the big batch of June placements went live, visibility across all prompts also climbed from about 20% to 23% (a 15% relative increase), landing right in that same two-to-three-week window.
In the broader competitive snapshot across non-branded commercial-contractor prompts, ServiceTitan sits first among tracked brands at 26% presence, ahead of competitors like BuildOps (21%), Jobber (8%), SimPRO (6%), and Housecall Pro (4%). In a category where every player is working the same AI surfaces, it consistently led the tracked field across the window.
April–May: Baseline tracking. ServiceTitan present but often mid-answer on commercial-contractor questions.
Late May–June: First placements go live and begin to index. Position starts shifting toward the top slot.
June onward: The commercial-contractor placements cluster and index. On the 50-plus-technician question, presence firms to a steady 75%-plus and citations climb following the placement cluster by about the two-to-three weeks these sources take to be absorbed.
Ongoing: July placements still maturing; results expected to keep building.
Focus on the right questions first. The work began with identifying the exact non-branded questions ServiceTitan's enterprise buyers ask AI, then placing into the independent sources those answers are built from. That's what drove the result: on the enterprise 50-plus-technician question, citations rose 250% as those specific sources were absorbed.
Showing up higher can matter more than showing up more. ServiceTitan moved from roughly fourth in the answer to roughly second - a 50% position gain - without needing to appear in more answers overall. A brand named second on a buyer's shortlist gets weighed differently than one buried at the bottom.
A new category is winnable, even as an established name against native rivals. Being known in the trades didn't hand ServiceTitan the commercial-contractor answer, but strategic placement in the sources AI cites got it to first among tracked brands at 26% presence, ahead of BuildOps, a rival built specifically for that buyer. Reputation doesn't transfer to an adjacent category on its own; targeted coverage in the right sources is how you break in and stay ahead.
Every percentage here is a relative change, not a percentage-point one. There was no control group, and ServiceTitan was running its own content and SEO work at the same time as Noble's placements, so these results show the two moving together, not proof that one caused the other. This work is also ongoing.
For the full methodology and raw data, contact Noble.
GET DISCOVERED ON
