Veeva won the CRM. The fight that actually matters hasn't started yet.

Everyone is writing migration handbooks for the Veeva and Salesforce divorce. Most of them are answering the wrong question.

For the better part of two years, the life sciences software conversation has had one main character: the end of the Veeva and Salesforce partnership. If you work anywhere near commercial pharma technology, you have seen the handbooks, the "ten things to know before you migrate" posts, the consultancy webinars promising a calm path through the storm. All of it treats the same event as the whole story. Veeva is leaving the Salesforce platform, you have to pick a CRM, here is how to plan the data migration.

That is a real decision and a genuinely painful one for a lot of organisations. It is also, I think, the least interesting part of what is happening. The CRM question is largely settled, the dust is landing roughly where you would expect, and the more important contest has barely begun. So let me put a stake in the ground.

I do not think this stops at CRM. I think the Salesforce ecosystem is already moving into the part of the market everyone assumes is untouchable: the regulated content and Medical Affairs stack. Vault. PromoMats. Publications. Medical information. MLR review and approval. The layer that has been Veeva's real moat all along. And the reason this matters is that it is not a prediction about something that might happen in three years. It has started, and most people are looking the other way.

What actually happened, for anyone keeping score

The shorthand version most people carry around is slightly wrong, so it is worth being precise.

Veeva's CRM was never independent of Salesforce. It was built on top of the Salesforce platform under a long-running agreement, extended in 2014 to run through 2025. In December 2022, Veeva announced it would not renew, and would instead move its CRM onto its own Vault platform. The formal separation lands in September 2025, with a wind-down period for legacy customers that Veeva has since pulled forward to end of support around 2029. So the "split" was Veeva walking away from a platform it had outgrown, not Salesforce being pushed out of a market it owned.

Salesforce did not sit still. It launched Life Sciences Cloud, built in partnership with IQVIA and seeded from IQVIA's OCE CRM, and has since layered its agentic AI work on top under the Agentforce banner. This is not vapourware. By the end of 2025 it had signed north of forty customers, including names that make the point on their own: AstraZeneca, Novartis, Chiesi, Takeda, Pfizer.

The other side of the ledger matters just as much. The majority of large pharma has stayed put. Veeva's Vault CRM has reached roughly 140 live customers, with around ten of the top twenty biopharmas committed and management expecting that number to climb. Veeva has guided to fiscal 2027 revenue comfortably ahead of consensus. Whatever you read into the defections, this is not a company in retreat.

One number worth correcting, because it gets repeated lazily: Veeva does not hold 98% of pharma cloud. Its CRM share sits closer to 80%, and the more telling figure is that it reportedly serves 47 of the top 50 pharmaceutical companies. That is dominance, but it is dominance with seams, and the seams are where this gets interesting.

CRM was always the soft target

If you want to understand why I think the regulated content layer is the real battleground, start by asking why Salesforce went after CRM first.

CRM is the part of the Veeva estate that looks most like generic enterprise software. Field force engagement, call planning, multichannel outreach, sample management. It is heavily regulated at the edges, but at its core it is a sales engagement system, and sales engagement is precisely what Salesforce has spent twenty-five years and a great deal of money becoming the best in the world at. Coming at CRM, Salesforce was attacking from strength into a category it understands better than almost anyone. Partnering with IQVIA bought it the pharma-specific data and domain credibility it lacked. That is a sensible first move, and it has worked well enough to make pharma technology leaders genuinely reconsider their assumptions.

But CRM was never Veeva's deepest moat. The moat is Vault, the regulated content and document platform that sits under PromoMats, RIM and regulatory submissions, Quality, Clinical, Safety, and Medical. This is where the switching costs become brutal: validated environments, 21 CFR Part 11 audit trails, electronic signatures, controlled document lifecycles, and a decade of accumulated configuration that maps directly onto how a company gets its product approved and keeps it compliant.

The conventional wisdom says that moat is unassailable, that Salesforce will bank its CRM wins and leave the regulated stack alone. I think the conventional wisdom is already wrong, and the evidence is sitting on Salesforce's own partner page.

The boundary is already moving, just not where people are looking

Here is the part the migration handbooks miss entirely.

Salesforce does not need to build a Medical Affairs suite itself. Its platform is a place other people build on, and they already have. Mavens, now part of Komodo Health, ships a Scientific Publications Cloud, a Medical Information Cloud, and a Research and Grants Cloud, all built natively on Salesforce. Read those product names again. Publications planning. Medical information. Grants and investigator-initiated trials. That is not CRM. That is the heart of Medical Affairs and Medical Communications, the exact territory people assume sits safely outside Salesforce's reach.

And it is not a skunkworks experiment. In May 2025 Salesforce formalised the whole thing with a Life Sciences Partner Network, naming a coalition of agencies, consultancies and ISVs, Mavens among them, explicitly chartered to move customers off legacy platforms and onto Life Sciences Cloud. Salesforce is not just selling a CRM. It is assembling an ecosystem whose stated purpose is to colonise the surrounding workflows.

This is the structural point that the CRM-versus-CRM framing obscures. Veeva built a walled garden. Beautifully built, deep, compliant, but a garden with walls, where every application is Veeva's and the integration story is "stay inside." Salesforce is the opposite shape. It is a platform, and platforms grow at the edges through people who are not the platform owner. The Medical Information Cloud from Mavens is already a pre-validated, auditable, Salesforce-powered content management system with Agentforce sitting on top of it. That sentence alone should end the argument that regulated content is somewhere Salesforce cannot go. It is already there, wearing a partner's badge.

So the question is no longer whether the Salesforce ecosystem can operate in Medical Communications. It demonstrably can and does. The question is how deep it goes, how fast, and whether Salesforce ever decides to pull these capabilities first-party.

The wedge is MLR, and the workflow is the easy bit

If you want to know where this contest gets decided, watch medical, legal, and regulatory review.

MLR is the chokepoint of pharma content. Every promotional asset, every claim, every piece of supporting material runs through it, and it is slow, expensive, and universally hated. It is also, at its core, exactly what a colleague put to me recently: an asset overlaid with a workflow. A document, some metadata, a routing and approval sequence, an audit trail, a sign-off. And once you frame it that way, the idea that this is beyond Salesforce falls apart, because configurable workflow with full auditability is not a frontier for Salesforce. It is the home turf. Flow, approval processes, electronic sign-off and audit history are the things the platform has done at enterprise scale for two decades. The Mavens medical information product already proves regulated content review runs on this substrate today.

So I will concede the point plainly, because it is correct. The MLR workflow itself is well within Salesforce's reach. If anything it plays to its strengths.

Where I would add a caveat is on the word "asset," because that is where the real difficulty hides and where the lazy version of this argument comes unstuck. The asset in PromoMats is not a generic file with an approval flow stapled to it. The domain value, the thing a decade of Veeva engineering actually bought, is the claims and references graph. Every promotional claim linked to the reference that substantiates it. Modular content components reused across markets and channels. Version lineage that holds up under inspection. Annotation and review layered directly onto the rendered asset rather than in a comment thread beside it. That is a specialised content-plus-linked-data problem with a purpose-built review experience on top, and it is not something you get for free from Salesforce Files and a flow.

That is the honest shape of it. The workflow is buildable, trivially so by Salesforce standards. The asset graph is buildable too, but it is real work and real domain depth, and it is the part that separates a credible PromoMats challenger from a glorified document approval app. The genuine moat was never the routing. It is the depth of the content model and the connective tissue that links it to the rest of the regulated estate.

Now add AI, because that is what makes this urgent rather than theoretical. Veeva has put agents straight into PromoMats, including a Quick Check that screens content against compliance and editorial guidelines before review even begins. An early user described it as moving them towards MLR becoming nearly touch-free. Sit with that phrase. Nearly touch-free MLR is not a feature, it is a reason to re-platform. And the value of that automation depends entirely on its grip on the underlying asset graph, the guidelines, the references and the workflow. Whoever owns that layer owns the AI that sits on it. Veeva is racing to make that grip native to Vault. The Salesforce ecosystem, with Agentforce and a head start in agentic orchestration, has every reason to argue the intelligence should live where the engagement data and an increasing slice of the Medical Affairs workflow already do.

The honest case against my own argument

I would not trust this piece if it did not include the reasons I might be wrong, so here they are.

The asset-graph depth I just described is exactly why Veeva's moat is not the kind of thing money alone solves. Building a credible claims and references model, validated to GxP standards, integrated across clinical, regulatory, quality and safety, is the work of years, and Veeva has been compounding it for over a decade. The connective tissue is the real lock-in: claims and content that flow between PromoMats, RIM and submissions, so that replacing one piece means unpicking all of it. That is why most of the top twenty are staying.

There is also a difference between ecosystem and commitment. What is live today in Medical Affairs on Salesforce is largely partner-led, ISVs like Mavens building on the platform, not Salesforce itself betting its roadmap on regulated content. Partner-led momentum is real, but it is a different and softer thing than a first-party Salesforce push, and it can stall, get acquired into irrelevance, or simply never reach the depth that validated submissions-grade content demands. Salesforce may decide the regulated content layer is a fine place for partners to play and a poor use of its own engineering, bank its CRM gains, and leave the deepest part of the stack to Veeva. If that is how it plays out, the boundary creep I am describing stays shallow, and Veeva's core holds.

So this is a genuine bet, not a certainty. I think the platform logic, the partner ecosystem, and the AI economics all point the same way. I accept that the deepest regulated content moat might hold regardless.

Where this leaves the rest of us

I have a stake here, so I will name it. At Nine Labs we build software for Medical Affairs teams, in the scientific communications and planning space, which sits right next to this fight. Watching two giants redraw the platform boundaries is not an academic exercise for anyone building in the regulated content world. When the foundations move, everything built on or beside them has to move too.

That is the part the handbooks miss. They frame this as a procurement decision: pick a CRM, plan the migration, mind the compliance gaps. It is bigger than that. We are watching the boundary lines of the entire life sciences software stack get redrawn, and the most consequential lines, the ones around regulated content, publications, medical information and AI-driven review, are already being crossed by a Salesforce ecosystem that the CRM story tells you to ignore.

My bet is that the next two years are defined not by who won the CRM, but by how far the Salesforce platform creeps into Medical Affairs, and whether Veeva's content moat is deep enough to hold when the attack comes through the ecosystem rather than head-on. I think CRM was the doorway, not the destination.