Veeva Is Rebuilding Its Own Monopoly

Veeva reported fiscal second-quarter results after the close on August 26, and the headline numbers were clean: revenue of $928 million, up 18% from a year earlier, adjusted earnings of $2.35 a share against the roughly $2.22 analysts expected, and a raised full-year outlook to about $3.68–3.69 billion. The stock traded up close to 9% after hours. None of that is the story.

Two quieter statements on the call are. Management said its end state in pharma CRM is holding over 70% market share. And it said the first Falcon go-lives arrive this year. Those are strange things for a company this dominant to need to say out loud. A business with more than 70% of a market doesn't usually announce its intention to keep it. It announces a plan to rebuild the product it has 70% of, and then quietly does it. That is what is happening, and the quarter handed investors a scoreboard to watch.

(Veeva's fiscal year ends January 31, so this quarter covered the three months through July, 2026.)

Veeva makes software for drug companies. Its best-known product, CRM, runs a drugmaker's commercial operation: which reps called on which doctors, what was said, what samples were left behind, all with the audit trail regulators require. The company was founded in 2007 on a strange plan — build on Salesforce, the generic cloud CRM, but customize it so completely for life sciences that it stopped being generic. It worked. Roughly 80% of pharma field teams ran on the thing, and the rest of the business grew on top of it.

Then in 2022 the host decided it wanted the life-sciences market itself, and the fifteen-year marriage ended. Veeva's answer was not to fight over terms. It would leave Salesforce's platform entirely, rebuild its commercial products on its own Vault technology, run them on AWS, and move the whole customer base over. It stopped selling the old Veeva CRM — the best-selling product in its history — to new customers, and set support for it to end in December 2029. Every CRM customer now has to pick a new home: Veeva's new product, Salesforce's, or something else.

That forced choice was the bear case. A mandatory migration hands competitors an opening, and investors braced for attrition; Veeva's stock fell from above $300 to a 52-week low near $148.

The opposite has happened so far. More than 180 customers are now live on Vault CRM, the replacement, up from around 150 at the end of the first quarter. Management called it the best CRM quarter ever. Commercial subscriptions grew 13%, the segment everyone had worried about. Two of the largest drugmakers in the world, Lilly and Biogen, chose Vault CRM in the quarter. Twelve of the top 20 biopharmas have now committed, two are still deciding and due to pick by year-end, and Veeva expects to win back accounts that chose Salesforce's product, mostly in 2027–28 as the old one's clock runs out.

Why it's working is not mysterious. Migrations are disruptive and heavily regulated; a drug company doesn't casually change its commercial engine. And the alternatives are green — Veeva says Salesforce still has no customers live on its new architecture. Dominance buys time. That's the whole reason the rebuild is possible at all.

But the deeper question is why Veeva rebuilt at all, instead of collecting rent on a 70%-share product until it ran out. The honest answer is that the product was running out. CRM for pharma was built to track a specific machine: the sales rep calling on doctors, with everything organized around that call. That machine has spent years getting smaller and moving online. A tracker of a shrinking thing is a shrinking business, whatever its share.

So Veeva has quietly changed what it believes it sells. Not software that records work. Software and agents that do the work. That is Falcon. Announced in May, Falcon is a set of standard AI agents inside Veeva's clinical, regulatory, and safety applications — sorting trial master-file documents, drafting responses to health-authority letters, triaging incoming safety cases. Veeva calls it "agentic labor." Early-adopter availability is planned for November, five early adopters are in place, and the first go-lives are expected this year — one at a top-20 drugmaker.

The detail worth sitting with is the price: $50 per user per month.

Think about what that number concedes. Per-seat pricing makes sense when a person uses a tool. It is a strange way to price an agent that does the work itself. The model underneath only gets cheaper, and eventually a customer will notice it's paying far more than the compute. Veeva seems aware of the tension. Its Falcon product for reviewing marketing claims, bought in June, is priced per document and per safety case — by the outcome, not the seat. The company is running both pricing models at once because it doesn't yet know what the category is. That's what an early version of a real product looks like, and it's also what a pilot risk looks like.

Read this way, the quarter's real content has nothing to do with the beat. It is two scoreboards. One: how many of Veeva's roughly 770 commercial customers go live on Vault CRM, and whether the last two undecided top-20 accounts land this year. Two: whether Falcon goes from five early adopters to paying seats in 2027 — pulled by customers, or pushed by salespeople.

Those scoreboards matter more than the raised guidance, because the guidance still mostly measures the old world. In the meantime the company is a machine: around 45% adjusted operating margins, about $1.6 billion of annual operating cash flow, more than $6 billion of cash and investments. At roughly a $40 billion market cap before the after-hours pop — about 27 times this year's expected adjusted earnings — the market is already paying for the rebuild to work and for Falcon to become something. Cheap is not the word.

Financially, the question comes down to whether the company that owns the tracker of last decade's machine gets to own the labor that replaces it. Veeva has made its own progress legible: it reports customer live counts, named wins, and pilot conversions every quarter, which is what a company making something people want looks like. The same scoreboard tells you when to worry. If the migration finishes, Falcon stalls at pilots, and the $50-a-seat price starts sounding like a story rather than a product, a 27-times multiple quietly becomes a 15-times one.

The quarter told you the old product is fine and the market's central fear was wrong. The next year answers an older question that no earnings number can: whether the new thing gets pulled, or pushed.