Sponsored placements now sit inside ChatGPT's own answers, and we wanted to know how evenly that inventory is actually being bought. Using our own daily multi-engine tracking, which includes ChatGPT ad tracking, we scored ad coverage across five very different categories over a rolling month: two enterprise B2B software categories, two field-service and compliance software categories, and one hospitality vertical. The spread between the most and least monetised category is 40 times. Waste, dumpster, septic and scrap hauling software carried a sponsored placement in 77.8% of the answers we tracked. London hospitality, wine bars and restaurants carried one in 1.9%. That gap is not random. The categories getting hit hardest are the high-consideration B2B software categories where a single click is worth real money to an advertiser, and the category left almost entirely alone is the one where a click is worth almost nothing.
TL;DR
- A 40x spread: ad coverage ranges from 77.8% of tracked answers (waste and hauling software) down to 1.9% (London hospitality).
- The two most heavily monetised categories, waste and hauling software (77.8%) and entity resolution and identity data infrastructure (73.0%), are both high-consideration B2B purchase categories with real annual contract value, not consumer verticals.
- The same advertisers buy across categories. Expert Market appears in four of the six categories we sampled and dominates finance automation outright. Jobber runs an identical "Jobber vs. ServiceTitan" conquest ad in two separate categories.
- UK trades and gas compliance software went from zero ad-bearing answers in its first two tracked weeks to carrying 29 of the month's 60 ad-bearing answers in its final three days.
How did we measure ad coverage across five categories?
We ran our own daily multi-engine tracking, which includes ChatGPT ad tracking, against five category platform sets over a rolling month from 29 June to 29 July 2026. Waste, dumpster, septic and scrap hauling software covers 23 tracked platforms including CurbWaste, Housecall Pro, ServiceTitan, Jobber and Routeware. Entity resolution covers 11 tracked platforms including Senzing, Tamr, Quantexa and Reltio. UK trades job management and gas compliance software covers 27 tracked platforms including Payaca, Tradify, Commusoft and simPRO. Finance automation and London hospitality round out the set. Coverage, in this study, is the share of tracked answers carrying a sponsored placement alongside the organic AI answer, not any single advertiser's share of that inventory.
We also track a sixth category, team collaboration and messaging software, but its measurement window is shorter than the other five, so we are not publishing a coverage percentage for it here. Its advertiser mix, further down this piece, rests on 326 ad-bearing answers, a large enough sample to be worth reporting on its own terms.
Which categories does ChatGPT monetise hardest?
Read down this table and the shape is stark.
| Category | Ad-bearing answers | Total tracked answers | Coverage |
|---|---|---|---|
| Waste, dumpster, septic and scrap hauling software | 1,070 | 1,375 | 77.8% |
| Entity resolution and identity data infrastructure | 1,132 | 1,550 | 73.0% |
| Finance automation, treasury, AP and payroll | 114 | 1,617 | 7.1% |
| UK trades job management and gas compliance software | 60 | 1,650 | 3.6% |
| London hospitality, wine bars and restaurants | 11 | 585 | 1.9% |
| Category | Value |
|---|---|
| Waste, dumpster, septic and scrap hauling software (highlighted) | 77.8% |
| Entity resolution and identity data infrastructure | 73% |
| Finance automation, treasury, AP and payroll | 7.1% |
| UK trades job management and gas compliance software | 3.6% |
| London hospitality, wine bars and restaurants | 1.9% |
Waste and hauling software and entity resolution sit more than ten times more monetised than finance automation, and more than 20 times more monetised than UK trades, despite finance automation and UK trades both being B2B software categories in their own right. Hospitality brings up the rear by a wide margin. Barely one in fifty of the answers we tracked carried an ad at all.
Why are two B2B software categories so much more heavily monetised than the rest?
The two hardest hit categories share a structural feature the other three do not. A waste hauling operator buying route management software, and an enterprise buying an entity resolution platform to clean up its customer data, are both making purchases with real annual contract value, exactly the kind of lead worth a serious bid in an ad auction. Finance automation and UK trades compliance software are B2B categories too, but their coverage, 7.1% and 3.6%, sits far closer to hospitality's 1.9% than to the top two. London hospitality has almost no advertiser willing to pay for the inventory at all: one restaurant bidding on Canary Wharf dining queries, surrounded by unrelated network fill like credit card rewards and travel booking ads. A one-off dinner booking simply is not worth bidding for the way an annual software contract is. We have written before about how little AI visibility this vertical gets organically too, in why your restaurant is invisible on ChatGPT: the same category competitors will not pay to advertise into is also one most restaurants are not showing up in for free.
Which advertisers buy across more than one category?
The same names keep showing up. In the answers we sampled from each category, five advertisers appeared in more than one vertical.
| Advertiser | Categories it appeared in (sampled) | Sampled appearances |
|---|---|---|
| Expert Market | Finance automation, UK trades, team collaboration, waste and hauling | 15 |
| Jobber | Waste and hauling, UK trades, team collaboration | 11 |
| ZoomInfo | Entity resolution, waste and hauling | 5 |
| Workato | Waste and hauling, finance automation | 2 |
| GlossGenius | Waste and hauling, team collaboration | 2 |
| Category | Value |
|---|---|
| Finance automation, treasury, AP and payroll (highlighted) | 7appearances |
| UK trades job management and gas compliance software | 4appearances |
| Team collaboration and messaging software | 2appearances |
| Waste, dumpster, septic and scrap hauling software | 2appearances |
Expert Market is the widest buyer of the set, showing up in four of the six categories we sampled and dominating finance automation outright with 7 of the 22 sampled ads there. Expert Market runs a lead generation model (quotes.expertmarket.co.uk), not a single product, which fits the pattern: it treats several adjacent B2B categories as one addressable audience rather than competing inside a single vertical. Talroo shows the same shape inside waste and hauling, running a "Talroo Connect for Driver Hiring" ad that targets the labour market around the category, not the software itself. ZoomInfo, a data platform rather than a lead generation network, shows up in both entity resolution and waste and hauling. Jobber, the field service platform, is the most consistent single product advertiser across categories, and its creative tells its own story.
What does the conquest creative actually look like?
Jobber runs the same competitor attack ad in two separate categories. In waste and hauling, the verbatim ad title is "Jobber vs. ServiceTitan," with the line "The easier ServiceTitan alternative. Try free." The identical "Jobber vs. ServiceTitan" title appears again inside UK trades and gas compliance software, a different category and a different platform set, run by the same advertiser attacking the same named competitor. Jobber also runs a non-conquest creative, "Jobber for Service Pros," offering "Quote, schedule and invoice. Get 40% off today," so the conquest ad is a deliberate second creative, not the only one in rotation. We saw direct conquest evidence elsewhere too: a prompt asking for an alternative to a named platform in waste and hauling served both a Jobber ad and a Grid AI Labs ad at once, meaning the switching-intent query itself is being bid on by more than one advertiser.
This matters for any brand that already ranks first organically in its category. Ranking first no longer means owning the answer. In waste and hauling, where 77.8% of tracked answers carry an ad, a brand named as the top organic recommendation can sit beside a paid placement from a direct competitor, running a "the easier alternative" pitch aimed straight at switching intent, in the same answer, at the same time.
How fast can a category's ad load move?
We are reporting what we observed, not why it happened. UK trades and gas compliance software carried zero ad-bearing answers across its first two tracked weeks. The first ad appeared on 7 July. Nothing followed between 8 and 16 July. From 17 July onward, ads appeared continuously and the pace accelerated: the final three days of the tracking window carried 29 of the month's 60 ad-bearing answers, roughly half the month's total inventory landing in three days.
| Metric | Value | Caption |
|---|---|---|
| Ad-bearing answers, first two tracked weeks | 0 | |
| Ad-bearing answers in the final three days | 29 of 60 | |
| Category coverage across the full window | 3.6% |
The other four categories moved differently. Waste and hauling and entity resolution both carried ads continuously across the whole window, with entity resolution's presence observed as far back as 26 May. Finance automation had one early placement on 5 July, then settled into steady coverage from 17 July, without the same step change UK trades showed. Hospitality had a single placement on 6 July, then ten placements landed in the window's final two days, a smaller-scale version of the same late acceleration pattern seen in UK trades.
What does this mean if you already rank first organically?
A brand that has done the work to rank first organically in a category no longer controls the full answer once that category starts carrying ads. The categories getting monetised hardest right now, entity resolution and waste and hauling, are exactly the categories where a competitor's paid placement is most likely to sit alongside your own citation, and the advertiser buying that placement may be running a conquest ad naming you directly. We go deep on the single hardest-hit B2B category in entity resolution's ad coverage and advertiser mix: every ad-bearing prompt we sampled there was unbranded and informational, meaning buyers are not waiting for someone to search a competitor's name, they are buying the research question itself.
FAQ
Which categories carry the most AI ads right now? Waste, dumpster, septic and scrap hauling software leads at 77.8% of tracked answers carrying a sponsored placement, followed closely by entity resolution and identity data infrastructure at 73.0%. Finance automation sits at 7.1%, UK trades at 3.6%, and London hospitality at 1.9%, a 40x spread from top to bottom.
Why do B2B software categories get monetised harder than consumer categories? The two most heavily monetised categories we measured, waste and hauling software and entity resolution, both involve high-consideration purchases with real annual contract value, which makes a single click worth bidding for. London hospitality, where a booking is a one-off, low-value decision, carries almost no paid inventory by comparison, at 1.9% coverage.
Do the same advertisers buy across multiple categories? Yes. In the answers we sampled, Expert Market appeared in four of the six categories we tracked and dominated finance automation outright. Jobber appeared in three categories and ran an identical "Jobber vs. ServiceTitan" conquest ad in two of them, waste and hauling and UK trades.
What does a conquest ad against a named competitor actually look like inside ChatGPT? In our tracking, Jobber's ad title reads "Jobber vs. ServiceTitan" with the line "The easier ServiceTitan alternative. Try free," appearing inside answers in both the waste and hauling and UK trades categories. We also observed a single switching intent prompt in waste and hauling serve ads from two different advertisers, Jobber and Grid AI Labs, at once.
Can a category's ad coverage change quickly? Yes. UK trades and gas compliance software carried zero ad-bearing answers in its first two tracked weeks, then carried 29 of the month's 60 total ad-bearing answers in its final three days. We report this as an observation from our tracking, not as evidence of any platform decision behind it.
Disclosure: Schmitdy is our own AI search service at AI Heroes, and this research comes from the same daily multi-engine tracking, which includes ChatGPT ad tracking, that we run for the categories we work in. We are publishing it as category-level research across the named platform sets above, not as a report on any single client, because a cross-category read of how AI answer inventory is actually being bought did not exist yet. For the deepest single-category read in this dataset, see entity resolution's ad coverage and advertiser mix. If you want to see whether your own category carries paid placements today, the free AI search audit shows it.
