On May 19, 2026, Salesforce said its Agentforce Life Sciences product was running inside more than 140 life sciences organizations, naming Novartis, AstraZeneca, Moderna, Pfizer, Takeda, AbbVie, and Merck Animal Health as customers (Salesforce). Five weeks later, on June 23, Veeva Systems, the vertical software company that has run life sciences data and regulatory workflows for two decades, answered. Veeva acquired Copli, maker of agentic medical, legal, and regulatory review software, and relaunched it as Veeva Falcon MLR, claiming the product could eliminate 70% or more of manual MLR review labor within five years (Veeva Systems Investor Relations).
That is the shape of the fight now defining vertical software. A horizontal platform with enormous distribution is naming a vertical incumbent's own customers in a press release. The incumbent's response is not a roadmap promise. It is an acquisition, a rebrand, and a specific labor-reduction number, published within weeks.
The acquire, don't build, pattern
Veeva did not improvise this move, and it is not alone. Vertical SaaS companies are converging on the same defense: buy agentic AI capability rather than build it from scratch or wait for a horizontal platform to take the workflow first. On January 20, 2026, Procore Technologies acquired Datagrid, an agentic AI platform built specifically for construction data, saying the deal would accelerate its AI strategy rather than build the capability internally or cede it to horizontal vendors (Procore / Business Wire). Four months later, on May 21, Procore shipped five AI agents, Deep Search, Submittals, RFI, Daily Log, and Contract Review, built on Datagrid's multimodal index and embedded directly inside the platform construction teams already use every day (Procore / Business Wire).
The logic is the same in both cases. Veeva and Procore did not need better AI research. They needed the agent inside their existing data and workflow before a customer had a reason to look elsewhere. An acquisition compresses that timeline from years to months.
Why the clock is running
Gartner forecast that 40% of enterprise applications will have task-specific AI agents embedded by the end of 2026, up from less than 5% in 2025 (Gartner). That forecast is the pressure behind every acquisition above. A vertical SaaS company that has not embedded an agent by the time its category crosses that threshold is competing on a feature set customers already consider baseline, not differentiated.
A horizontal platform naming your customers in its own press release is not a warning shot. It is a sales call already in progress.
The other path: AI natives that skip the incumbent entirely
Not every vertical fight runs through an incumbent's acquisition budget. Legal AI company Harvey closed a $200 million growth round co-led by GIC and Sequoia Capital on March 25, 2026, at an $11 billion valuation, up from $8 billion just months earlier (CNBC). Research firm Sacra estimated Harvey's annualized recurring revenue reached $300 million by its June 2026 update, up from $195 million at the end of 2025, a third-party estimate rather than a company-confirmed figure (Sacra). Sierra, the enterprise AI agent company founded by Bret Taylor, raised $950 million in a round led by Tiger Global and GV on May 4, 2026, valuing it above $15 billion and giving it more than $1 billion in capital to compete for enterprise customer experience workflows against horizontal platforms (TechCrunch).
Harvey and Sierra never had a legacy platform to defend, so they built the vertical data advantage directly and funded it at a scale most incumbents cannot match from cash flow alone. Harvey's ARR growth suggests the approach can outpace any incumbent's response time. That is the second track running parallel to Veeva and Procore: the incumbent's defense is acquisition, the native challenger's offense is capital.
What to do this week
- List every workflow in your product where a horizontal platform (Salesforce, Microsoft, Google) could plausibly ship a task-specific agent within two quarters.
- For each one, decide now whether you build, buy, or accept the loss. Veeva and Procore both chose buy, and both published a public number (a percentage, a named agent set) within months of the acquisition closing.
- If a horizontal platform has already named your customers in a press release, you are past the planning stage. Match Veeva's timeline: acquisition to public relaunch in under six months.