In a remarkable 30-day period, three B2B software companies—Fin (Intercom), Cognite, and MaintainX—were each acquired for approximately $3 billion by Salesforce, Schneider Electric, and Autodesk respectively. At first glance, these companies operate in entirely different sectors: customer service, industrial data platforms, and facilities maintenance. Yet every deal shared the same core asset: proprietary data that makes AI effective in a specific domain.
This wave of acquisitions reveals a fundamental shift in how investors and acquirers value B2B software companies. The message is clear—domain-specific data has become the primary AI moat, and buyers will pay extraordinary premiums for companies that own it.
The Data Behind the Deals
Despite operating in distinct verticals, these three acquisitions followed a consistent pattern. Each company controlled unique datasets that train or contextualize AI for their specific category, and each demonstrated rapid monetization of AI products built on that data foundation.
Fin's story is particularly instructive. The company's AI agent line grew 350% year-over-year to reach $100M in annual recurring revenue—even as legacy revenue remained flat. Acquirers paid premiums that tracked growth velocity of the AI product lines rather than blended multiples across the entire business. This wasn't about buying mature cash flow; it was about securing access to proprietary data and the AI monetization engine built on top of it.
The premium valuations—30x+ for fast-growing AI segments—signal that investors and strategic buyers are willing to overlook stagnant legacy businesses when the AI component demonstrates both unique data assets and proven revenue growth.
Why Data Is the Real Moat
Model architecture alone doesn't create defensibility. Foundation models are increasingly commoditized, and any startup can call an API. The defensible advantage lies in owning the training data and contextual information that makes AI accurate, relevant, and useful within a specific domain.
Consider what each acquirer actually purchased:
- Salesforce acquired Fin's customer service interaction data and conversation patterns
- Schneider Electric acquired Cognite's industrial operations and IoT sensor data
- Autodesk acquired MaintainX's facilities maintenance records and workflow data
These datasets cannot be replicated by competitors. They're built up over years of customer relationships, integrations, and operational history. When paired with AI models, they create compound value that grows more defensible over time—the more customers use the product, the better the data, the smarter the AI, the more valuable the offering.
What This Means for Founders Building Now
If you're building in a vertical market, your path to a premium exit depends on controlling the data layer for your category. This insight has immediate implications for product strategy and go-to-market execution.
First, identify what unique data your product can capture that competitors cannot easily replicate. This might be workflow data, decision data, outcome data, or domain-specific interaction patterns. The key is exclusivity and relevance to AI applications in your space.
Second, prove AI monetization early. Don't wait until you have perfect data coverage to launch AI features. Fin demonstrated that a fast-growing AI product line at $100M ARR can drive a $3B acquisition even when core revenue is flat. Investors want evidence that your data translates into AI products customers will pay for.
Third, focus on growth velocity in your AI segments over blended metrics. Acquirers are paying for future potential, and 350% year-over-year growth in an AI product line signals that potential more clearly than steady 20% growth across legacy offerings.
Key Takeaways
- Three $3B acquisitions in 30 days all centered on proprietary, domain-specific data for AI
- Fin grew its AI agent line 350% YoY to $100M ARR despite flat legacy revenue, proving data + AI monetization drives premium valuations
- Model architecture is commoditized; defensible data in a specific vertical is the real AI moat
- Acquirers paid 30x+ multiples for fast-growing AI segments, focusing on growth velocity over blended company metrics
- For founders: own the data layer in your category and prove AI monetization early to unlock premium exit potential
The evidence is conclusive—investors and acquirers are hunting for companies that control valuable, domain-specific data. If you're building in a vertical, your strategic priority should be capturing unique datasets and demonstrating that AI products built on that data can generate real revenue growth. The market will reward you accordingly.