ServiceTitan财报拆解：AI转型期的收入阵痛与GTM取舍

“Vertical SaaS for The Traders” leader ServiceTitan reported its fiscal Q2 2027 on September 8 (quarter ended July 31, 2026), and on paper it was a good quarter. Revenue of $292.8M, up 21%, ahead of their own 18% guide and ahead of consensus. Non-GAAP EPS of $0.40 against $0.36 expected. Operating margin up 310 basis points. Record free cash flow. Net dollar retention still above 110%. They even raised the full-year revenue guide, by $4M.

The stock fell 30% the next day. More than $2B of market cap gone in a session. By Friday it had hit a 52-week low of $54.16, down roughly 52% over the past year.

The quarter didn’t cause that. The back half did: revenue guided to roughly 15% growth, against 25% a year ago. Four quarters of deceleration, and a Q3 revenue guide that comes in below Q2 in absolute dollars.

Grow or die.

What happened:

• Revenue +21%, a beat on both revenue and EPS, and the stock still lost 30% in a day
• Q3 revenue guided to $285M-$287M, below Q2’s $292.8M, with the back half implying ~15% growth
• Gross transaction volume growth fell from 23% in Q1 to 17%, on soft HVAC lead volume in May and June
• Max, their agentic operating system, suppresses revenue in the near term by design: they don’t bill the first quarter of the contract
• They deferred expansion into new trades to concentrate resources on Max
• Non-GAAP operating margin hit 15.2% and free cash flow hit $50.5M, and the market paid nothing for either

#1. They Beat Revenue by $8M and EPS by $0.04, and The Stock fell 30% the Next day. Why? Weak Projected Forward Growth

The quarter itself:

• Total revenue: $292.8M, up 21%
• Platform revenue: $284.5M, up 22%
• Subscription: $212.4M, up 22%
• Usage: $72.1M, up 24%
• Non-GAAP operating income: $44.4M, a 15.2% margin, up from 12.1%
• GAAP operating loss: $27.6M, improved from $34.8M
• Free cash flow: $50.5M, up 47%
• Net dollar retention: above 110%
• Cash: $479.5M

They beat their own revenue guide by about $8M. They beat the Street. Margins expanded. Cash generation nearly doubled off a small base. Every metric a CFO would put on a board slide moved the right way.

Then the stock closed down $24.46, or 30%, the following session. Seven firms cut price targets within 24 hours: TD Cowen from $125 to $100, Wells Fargo to $105, Piper Sandler and KeyBanc to $110, BMO from $103 to $90, Canaccord from $105 to $90. Almost all of them kept their Buy ratings while cutting the number by 20-30%. They didn’t downgrade the business. They repriced the forward growth rate, which is the only line in the release that got worse: 21% this quarter, roughly 15% guided for the next two.

#2. A year ago the back half grew 25%. Now it’s guided to 15%

Quarter by quarter:

The Q3 and Q4 FY27 figures are arithmetic off the company’s own guidance table, and management confirmed it on the call: the back half is planned at mid-teens growth. H1 grew 22.7%. H2 is planned at roughly 15%.

Q3 revenue is guided below Q2 in absolute dollars. $292.8M reported, $286M guided. A software company at a $1.17B run rate telling the market its next quarter is smaller than its last one is a hard thing to explain, even when there’s a real reason for it.

The margin steps down with it. Non-GAAP operating income is guided to $29M-$30M in Q3 on $286M, about a 10.3% margin, versus 15.2% in Q2. Growth down and margin down in the same guide. On a Rule of 40 basis the back half runs around 27, against 36 in the quarter they just reported and 37 a year ago.

The full-year guide going up $4M was technically true and completely irrelevant. Look at how little that annual number has told you all year: ServiceTitan set FY27 guidance at $1.11B-$1.12B last March, raised it twice, and now sits at $1.139B-$1.144B. Over that same stretch the stock is down more than 50%. The market has been trading the slope, not the total.

#3. GTV growth dropped to 17%, and that lands in the P&L the same quarter

ServiceTitan is priced off its customers’ volume. Gross transaction volume, the total dollars their contractors invoice through the platform, was $26.8B in the quarter, up 17%. Revenue as a percentage of GTV is about 1.09%, up from 1.06% a year ago. They take roughly a penny of every dollar their customers bill.

That’s a great model on the way up and an exposed one when the end market wobbles. GTV growth went from 23% in Q1 to 17% in Q2. Management’s explanation was specific: lead and job volume softened across the industry in May and June, concentrated in HVAC, customers didn’t offset lower volume with bigger tickets, an early cooling season likely pulled some demand into Q1, and the observed July 3 holiday behaved like a weekend day. Leads improved through the quarter and stabilized in July. Management pointedly did not declare the slowdown over.

Revenue still grew 21% on GTV of 17%, so they out-monetized their own volume by four points. Usage revenue grew 24%, well ahead of GTV. Monetization did its job and still landed the quarter at 21% instead of 25%, because six points of HVAC volume showed up in the same three months it happened. A seat contract would have carried that softness into next year’s renewal instead.

#4. Their AI Max doesn’t bill for the first quarter of the contract

Max is ServiceTitan’s agentic operating system, and it’s the center of the strategy. Adoption is moving: just over 100 enrolled locations at the end of Q1, they exceeded their goal of doubling that in Q2, and they now expect more than 700 by the end of the fiscal year.

And it is costing them revenue right now. Two separate drags they disclosed:

• Max requires enough change management that, in their words, they “typically do not bill for the first quarter of a contract,” then ramp to full contract value over roughly the first year
• A revenue recognition timing difference between core and upsell creates a $2M-$3M subscription headwind over the remainder of the fiscal year

Together, roughly $4M-$5M of near-term revenue headwind from the mix shift. The faster Max sells, the more it suppresses reported revenue in the quarter it sells.

The logic is sound. If the product genuinely takes a contractor’s technician-to-admin ratio from 2:1 to 3:1, waiving the implementation quarter is a rational way to remove friction on a hard change-management sale. Early adopters are reportedly growing revenue 45%+.

It also means the AI transition shows up in the P&L as a growth headwind for four to six quarters before it shows up as expansion. BMO put the material benefit in FY28 with good execution. In 2021 the market funded that kind of gap. In 2026 it takes 30% out of the stock while you cross it.

If you’re planning an AI repackage that changes your billing model, price the trough it puts in your own revenue line before you commit to it. ServiceTitan’s trough is worth $4M-$5M a year against a $1.14B base, and it still cost them a third of the company.

#5. They deferred expansion into new trades to fund the AI bet

ServiceTitan is intentionally narrowing what it’s working on: deferring expansion into new commercial and residential exterior trades to concentrate resources on Max and on their internal “software factory” effort.

You can see the choice in the expense lines:

• R&D: $100.6M, up 38%, versus 21% revenue growth
• S&M: $77.0M, up 10.7%, half the rate of revenue growth
• G&A: $59.0M, down 7%

They shifted dollars from selling to building, and from new markets to the existing base. That is a defensible allocation. Max monetizes inside customers they already have, at a base with NRR over 110%, which is cheaper than buying new trades.

It also took two growth vectors out of the forecast at the exact moment the core growth rate was decelerating. Commercial and residential exteriors were the headroom in the model, and they moved to the right in the same quarter the base slowed to 21%.