CS Disco Grew 15% and Called It a Category — the Street Already Paid +52% for It
The best growth quarter in four years, a quarter-early EBITDA positive, and a new label — AI for litigators — sit on top of a 13%-growth guidance range and a competition question management keeps answering sideways
LAW · Earnings Call · 2026-11-04
The best quarter in four years — and the ceiling it quietly implies
CS DISCO (LAW) is a $367M micro-cap that has spent its public life going the wrong way: down 86% since its 2021 listing, 91% off the $65.88 peak. Then something turned. The stock is up 51.6% over the trailing 90 trading days, a one-segment melt-up into an earnings report that, for once, had a number to justify it. Q3 FY26 total revenue was $45.3M, up 15% year over year; software $38.6M (+13%), services $6.7M (+21%). Adjusted EBITDA came in positive a quarter ahead of the company's own Q4 target. Eric Friedrichsen did not undersell it: the quarter was, by his telling, “the best growth quarter that we have had in 4 years” — Eric Friedrichsen, Chief Executive Officer · 2026-11-04, accelerating from 13% in Q2. Small wonder the tape moved first. But here is the fine print the market tends to skip. All of those growth rates are adjusted — a $1.3M contingent case recognized in Q3 of last year inflates the comparison base. Include it and the numbers are 13% total, 11% software. Aaron Barfoot was explicit: “the revenue associated with the matter has been on our platform for more than 5 years and was contingent on the successful outcome of the case” — Aaron Barfoot, Chief Financial Officer · 2026-11-04. That makes contingent matter a fresh top-five keyword for the company and, simultaneously, a reminder that the headline beat is a few points cleaner than it looks. The filed fundamentals lag the call — the latest 10-Q (filed 2026-08-05) covers the quarter ended 2026-07-31, i.e. the $43M Q2, not the $45.3M just reported. Even so, they frame the re-rating. Price to revenue sits at 1.5x, down 81% in a year and 94% from the 2021 peak of 23.1x. At that multiple, a 15% growth quarter with breakeven EBITDA is a genuine option on something better — which is exactly how the stock traded.Cecilia grows up: from product to category
What is actually new here is less the revenue than the vocabulary. The company's top keyword this quarter is AI for litigators — a phrase absent from prior quarters and clearly a deliberate act of positioning: “DISCO is hyper focused on being the AI for litigators” — Eric Friedrichsen, Chief Executive Officer · 2026-11-04. The pitch is that litigation is not a feature you bolt onto a general legal AI stack; it is an architecture. That framing is doing real work in the sales motion, and it is being reinforced by three artifacts that show up as fresh keywords: Advanced research, AI review, and the Unified Litigation Solution now in pilot with several Am Law 50 firms. The adoption data genuinely moved. Advanced Research went generally available in Q3, grew to 91% of all Cecilia Q&A usage, and hit 97% in October; usage skews to the largest matters. Auto Review revenue grew more than 4x year over year, and one corporate customer reportedly mandated it across all outside counsel absent an exception. Large-matter count tripled on the new DISCO platform, which is now the default commercial model for every new and renewal contract. Customers spending over $100k in the last twelve months rose to 374, from 326 a year ago and 354 last quarter, representing $133M, or 78% of trailing revenue. That is a coherent machine. It is also a machine that was described, in almost the same words, three quarters ago — the unified platform was announced into pilot last May with an explicit instruction from the CFO: “I wouldn't build revenue from the solution into your models yet” — Aaron Barfoot · 2026-08-05. Nothing on this call changed that. The pilot expanded to a second cohort, a third is expected before year-end, and broad availability is still undefined.The keyword that quietly vanished
Management stopped saying some things. Wallet share — a top-30 staple through 20251–20253 and the structural core of the whole thesis — did not make this quarter's list, despite the CEO walking through 15–20% wallet penetration numbers in Q&A. Cecilia AI, the flagship brand that ranked first as recently as 20254, is gone as a keyword; the product has been absorbed into Advanced Research and the platform. Auto Review has been renamed in the discourse as generic AI review. Brand names fading into category language is usually a sign a company believes the category, not the SKU, is the moat. One oddity worth flagging: the two executives gave different adjusted EBITDA figures for the same quarter — Friedrichsen cited $79 thousand, Barfoot $100 thousand. Trivial in dollars, but on a call whose entire investment case rests on credibility of a profitability inflection, a nine-figure-in-thousands mismatch is not the detail to fumble.The tension: a 20% aspiration sold against 13% guidance
The recurring line is the 20%-plus grower claim, now in its fourth consecutive quarter of airtime. In February the CEO said it flatly:On Wednesday he repeated it — carefully hedged, not ready to call any specific quarter — while guiding FY26 to $176.1–178.1M against roughly $157M last year. That is 13%, not 20%. The gap is bridged by hope that DISCO platform's larger, longer matters compound, offsetting the lower one-time ingest fees management has now flagged for a third straight quarter as immaterial-to-date. The other recurring theme is competition, and the answer has not changed in a year. Asked whether cheaper frontier models commoditize first-pass review, Friedrichsen redirected entirely to the labor pool:I have said before that I believe that CS Disco, Inc. can be a 20% plus grower, and I am actually more optimistic than ever.
That is a strong market argument — the managed-review TAM is measured in billions and DISCO has never booked more than about $10M of human review — but it is also the same non-answer given in February, when he insisted “I have not heard of a single customer utilizing general AI or these frontier models for the eDiscovery process” — Eric Friedrichsen, Chief Executive Officer · 2026-02-25. Analysts have now asked a version of that question for three straight quarters. The deflections are consistent; the consistency is the tell. Margin structure backs the operating story at least: operating margin improved 9.6 points year over year to -21.4%, and free cash flow margin less stock comp improved to -16.9% from -30%-plus a year ago. For a company that used to lose a third of revenue on operations, that is the real change. The bottom line: DISCO put up its best growth in four years, reached breakeven early, and finally has a category story that is its own rather than borrowed from the legal-AI hype cycle. But the report also shows a business guiding to 13% while promising 20%, riding a single platform transition whose revenue mechanics (lower ingest fees, pilot-stage unified solution) remain deliberately unmodeled. The stock paid 52% in advance for the story. The next two quarters have to prove the guidance is the floor, not the ceiling.it is not the competition from a technology perspective that is what we are thinking a lot about. it is the human review. That is where the real opportunity is for services to become software.