The state of the ServiceTitan pricebook: what 27 real books show
Everyone has an opinion about pricebooks. We wanted numbers. So we measured what real ServiceTitan pricebooks actually look like inside, across a real sample, and the pattern turned out to be remarkably consistent from one shop to the next.
The short version: shops describe the work fine. What almost none of them have turned on is the selling structure ServiceTitan can prompt from. The fields are there. They are empty.
The numbers
| What we measured | Across 27 books, 65,819 active services |
|---|---|
| Services with a description | 100% |
| Services with no Recommendation set | 90.4% |
| Services with no Upgrade set | 95.1% |
| Jobs with no linked material | 64.3% |
| Books at a flat zero on recommendations | 13 of 27 (48%) |
Read those top two rows together. Every service is described, so the office did the work of writing the book. But the Recommendation and Upgrade fields, the ones ServiceTitan can surface to a tech as “offer this too” or “here is the better version,” sit empty on nine and nineteen out of twenty services. The platform is ready to prompt. There is nothing loaded for it to prompt with.
The material row is the one that costs the most quietly. When a service has no linked material, the parts it consumes do not roll into its cost, so margin is a guess and job costing never quite reconciles. Two thirds of jobs are in that state.
Why the pattern is so consistent
This is not a story about any one messy shop. The same shape shows up book after book, which tells you it is structural, not sloppiness. A pricebook gets stood up to bill work, the descriptions get written because you cannot invoice without them, and then the parts that make the book sell, the tiered options and the material links, get postponed because they compete with dispatch and payroll for someone’s time. So they wait. And they keep waiting.
That is not a knock on owners. It is the honest state of most books, and it is money the system could be prompting and is not.
How we measured it (so you can trust it)
We care that this holds up, because we are asking you to cite it.
- Real books, de-duplicated by ServiceTitan’s own IDs. Shops export their book many times. We fingerprinted each file by its set of active-service ServiceTitan IDs and clustered any two files sharing most of those IDs as the same account, keeping the largest export per account. So a shop’s dozen re-exports count as one book, not twelve.
- Active services only. ServiceTitan deactivates rather than deletes, so counting inactive rows would inflate everything. We count only affirmatively active services.
- Service-weighted percentages. A tiny book cannot swing the numbers; every percentage is across all 65,819 services, not an average of book averages.
- Cross-checked four ways. The original hand census, a filename de-dupe, a named-account de-dupe, and this full ID de-dupe all land on the same headline (roughly 90 / 95 / 64). When four methods agree, the number is real.
We publish the aggregate only. No client names, no prices, no per-book figures.
What it means for your shop
If your book looks like the average one in this sample, the opportunity is not “write more services.” You already did that. It is turning on the parts the platform is waiting to use:
- Recommendations and upgrades on the services that carry them, so a repair can become a choice. That is a 90%-empty field across real books. (How to build recommendations and upgrades in ServiceTitan.)
- Material links on the jobs that consume parts, so cost is real and margin stops being a guess. Two thirds of jobs are missing them. (Do you need to link materials in ServiceTitan?)
- Tiered good/better/best on your top services, structured so the tech presents instead of building a quote in the driveway. That is the heart of the pricebook that sells.
Want to see where your own book sits against this? The free Pricebook Health Audit reads your real ServiceTitan export in your browser and shows you your own version of these numbers in a few minutes. Your file never leaves your computer. If you would rather have it built for you, that is what our done-for-you pricebook builds are.
Cite this study
You are welcome to reference these figures. Please cite them as:
“The State of the ServiceTitan Pricebook,” Titan Tech Tools, 2026. Aggregate across 27 real ServiceTitan pricebooks (65,819 active services): 90% of services had no recommendation set, 95% no upgrade, and 64% of jobs had no linked material.
Link: https://titantechtools.com/resources/state-of-the-servicetitan-pricebook/
If you are writing something and want a specific cut of the data, get in touch and we will pull it.