ServiceTitan's AI Bet Gets Expensive: What a 30% Stock Drop Says — and Doesn't Say — About Agentic Software for the Trades

Status check — Sept. 22, 2026

ServiceTitan's most important product bet is now being examined from two very different directions. Contractors are being asked to adopt more automation through Max. Investors are asking how quickly that adoption converts into recognized revenue, durable growth and a larger economic relationship with the customer.

The second question became much louder after ServiceTitan shares fell 29.98% on Sept. 9, from $81.58 to $57.12. Dispatch previously examined the selloff itself and the roughly $2.36 billion one-day market-capitalization change. The new development is that multiple plaintiff-side securities firms have opened investigations and are soliciting affected investors.

That does not turn a volatile earnings reaction into evidence of misconduct. It does make the Max transition worth examining with more precision, because the facts ServiceTitan disclosed are unusually relevant to contractors buying agentic software: substantial change management, delayed billing during rollout, waived onboarding fees, a larger eventual subscription relationship and a strategic decision to put more investment behind AI.

What is established — and what is only alleged

• ServiceTitan reported Q2 FY2027 revenue of $292.8 million, up 21% year over year, and more than $50 million of non-GAAP free cash flow.
• ServiceTitan said it exceeded its goal of doubling Max locations during Q2 and now expects more than 700 enrolled Max locations by fiscal year-end.
• Prepared remarks state that Max requires substantial change management; ServiceTitan typically does not bill the first quarter of an upsell Max contract and ramps to full contract value over roughly the first year.
• ServiceTitan expects the changing mix toward Max and different revenue-recognition timing to create a $2 million to $3 million subscription-revenue headwind over the remainder of the fiscal year, plus roughly $2 million of professional-services pressure from waived onboarding fees.
• ServiceTitan's 10-Q says it elected to focus more tightly on existing commercial trades and residential roofing rather than continue planned expansion into additional commercial and broader residential-exterior categories at the same pace, in order to increase investment and attention on Max and internal AI.
• Private law firms are investigating whether prior disclosures were sufficiently transparent. Those allegations have not been proven, and the investigation announcements are not findings by a court, regulator or independent fact-finder.

Hagens Berman said its investigation is focused on whether ServiceTitan was sufficiently transparent about the financial ramifications of the Max push, including the timing difference between core subscription revenue and upsell products. Pomerantz announced a separate investigation concerning whether ServiceTitan and certain officers or directors may have engaged in unlawful business practices. Bleichmar Fonti & Auld said today that it is investigating statements related to Max and the company's expansion plans.

Those are the firms' allegations and theories, not Dispatch findings. Plaintiff-side law firms routinely announce investigations after sharp stock declines and invite investors to contact them. Some inquiries become lawsuits; others do not. Even a filed complaint would represent allegations until resolved.

The useful editorial question is therefore narrower: what did ServiceTitan itself disclose, and why did investors react so strongly?

The Max transition: company disclosures versus law-firm scrutiny

ServiceTitan's prepared remarks say Max requires substantial change management. That is more important to contractors than the stock chart.

Agentic software is not a normal feature upgrade when it changes who answers the phone, how leads are worked, how estimates are followed up, how dispatch decisions are made or how office employees spend their day. The software may be capable on day one while the organization takes months to redesign workflows, train staff, define escalation rules and build trust in the output.

ServiceTitan's billing model appears to acknowledge that reality. The company says it typically does not bill the first quarter of a Max upsell contract and then ramps toward full contract value over the first year or so. For a software company, that creates a lag between enrollment and reported revenue. For a contractor, it should be a warning against evaluating an AI project by activation date alone.

A 30% stock decline does not show that Max is failing with contractors. ServiceTitan reported the opposite on enrollment: Max locations grew faster than the company's target, and management raised its year-end enrollment expectation.

The decline also does not show that ServiceTitan is facing a liquidity crisis. The company reported more than $50 million of non-GAAP free cash flow in Q2 and positive non-GAAP operating income. Nor do private securities-law investigations prove that management misled investors.

What the decline does show is that public markets are sensitive to timing. A company can have strong product adoption and still be repriced if investors believe the path from adoption to reported growth is slower, more expensive or less predictable than expected. Contractors should recognize the same distinction inside their own P&L.

ServiceTitan says fully ramped Max customers roughly double subscription revenue relative to prior spend. That can be a good trade if the platform creates more than enough additional value to cover the higher cost. It can be a bad trade if the contractor buys automation without changing the operating system around it.

Every AI product should therefore have an economic hypothesis before rollout. An AI phone agent should increase booked revenue or lower cost per booked call. Sales automation should lift close rate, gross margin or salesperson capacity. Dispatch automation should create more productive technician hours. Estimate follow-up should recover otherwise lost jobs. Back-office agents should remove hours or increase throughput.

If the only measurable outcome is that the company now has more AI features, the business case is incomplete.

How to evaluate an agentic software rollout

What contractor owners should do before a major AI upgrade

• Write down the specific labor, revenue or gross-margin problem the agent is expected to solve before enabling it.
• Price the full first-year cost, including implementation time, integration work, staff training, usage charges and management attention.
• Set a 30-, 60- and 90-day adoption plan instead of treating activation as completion.
• Define a baseline for booking rate, close rate, jobs per tech, labor hours or another relevant metric before the rollout.
• Keep a human escalation path for customer-impacting decisions and audit the agent's claims, pricing language and handoffs.
• Review total software spend per technician and per completed job as more AI products move into premium tiers.
• Test data export and integration dependencies so a strategic vendor shift does not become an emergency migration.

ServiceTitan has made a clear strategic choice: concentrate more investment on Max and internal AI, deepen the product in core trade categories and accept some near-term revenue friction while customers work through a substantial operational transition.

Investors are now debating whether that transition was understood and communicated clearly enough. Private law firms are testing a legal theory around that question, but no public finding reviewed by Dispatch establishes wrongdoing.

For contractors, the lesson is more concrete. Agentic software is leaving the demo phase and entering the P&L. The products can be transformative, but the bill, the change-management work and the measurement discipline arrive with them. The standard should be the same whether the vendor is ServiceTitan, TradeVulcan or anyone else: prove the operating result.

Methodology

TradeVulcan Dispatch reviewed ServiceTitan's Sept. 8, 2026 earnings release, prepared remarks and Form 10-Q; compared those primary-source disclosures with Sept. 17 investigation announcements from Hagens Berman and Pomerantz and the Sept. 22 announcement from Bleichmar Fonti & Auld. Law-firm statements are identified as allegations or investigative theories and are not treated as findings of fact. Dispatch also cross-referenced its Sept. 8 earnings analysis and Sept. 10 stock-selloff analysis to avoid duplicating prior coverage. The hero is a real licensed Unsplash photograph used only as illustrative financial-market imagery.