The Great Rebundling: Why Vertical SaaS Is Becoming AI-First
For 20 years, SaaS unbundled industries. AI is rebundling them—around data, context, and workflow depth.
1. The End of the Horizontal Era
For the last two decades, software ate the world by slicing it into clean, horizontal layers. Sales tools. HR systems. CRMs. Every workflow got its own tab. The result was an explosion of innovation—and fragmentation.
But in 2025, something is shifting. The AI wave has exposed the limits of generic tooling. The value no longer lies in broad platforms with surface-level integrations; it lies in deep vertical stacks where data, decisioning, and automation live in the same loop. In short: AI doesn’t want APIs—it wants context.
2. AI Thrives Where Context Lives
Horizontal SaaS worked because humans could translate context. A sales rep could jump between Salesforce, Slack, and Excel and stitch together meaning. AI can’t. It performs best when it has domain-specific data, consistent feedback loops, and tightly coupled workflows.
This is why we’re seeing a rebundling: AI systems need vertical context to reason, automate, and act. Healthcare AI platforms that integrate patient data, billing, and compliance in one stack. Construction AI tools that own sensor data, inventory, and scheduling. Legal AI assistants that live inside matter management systems, not beside them.
The deeper the data roots, the smarter the automation.
3. From Tools to Systems of Intelligence
In every vertical, we’re moving from point solutions to what I call systems of intelligence:
First generation: digital record-keeping (ERP, CRM)
Second generation: workflow management (SaaS unbundling)
Third generation: decision automation (AI rebundling)
Each step pulls the software layer closer to the operating core of the industry. Instead of selling a tool, the new AI-first verticals become co-pilots of the business itself.
At Revolut, we saw this first-hand—AI wasn’t most valuable in the lab; it was valuable when plugged directly into payments, risk, and onboarding systems. The same lesson now applies across industries: AI becomes powerful when it’s operational, not ornamental.
4. Distribution, Not Differentiation, Wins the Market
Most founders over-index on model quality. But as the AI infrastructure layer commoditizes, the defensibility shifts to distribution—and distribution in vertical SaaS means data gravity and embeddedness.
Whoever controls the daily workflow wins the right to deploy intelligence. Once AI is inside a company’s heartbeat—billing, forecasting, inventory, or scheduling—it stops being an optional feature and becomes infrastructure.
That’s the new flywheel: distribution creates data, data trains models, models deepen product value, which reinforces distribution.
5. A New Design Language for AI-First SaaS
The design playbook is changing too. The next generation of vertical SaaS products will look less like dashboards and more like conversational, adaptive systems. Instead of clicking through menus, users will describe outcomes. Instead of integrations, they’ll get reasoning. The interface becomes invisible.
This isn’t UX minimalism; it’s cognitive leverage.
Every great AI-first SaaS company will need to master this trifecta:
Data proximity – be as close as possible to the raw, structured source of truth.
Workflow embedding – own the daily motion of how work gets done.
Adaptive interfaces – shift from command-based to conversation-based systems.
6. What This Means for Builders and Investors
For builders: stop thinking of AI as a feature layer. Build it into the bloodstream of your vertical. Find where decisions are made, where humans still bridge systems, and replace those seams with intelligence.
For investors: stop chasing general-purpose copilots. Look for founders who have insider-level empathy for an industry’s data stack and operational rhythm. The winning companies will feel more like domain-native operating systems than traditional SaaS.
The next era of software isn’t about serving users across industries. It’s about owning an industry end to end.
7. The Rebundled Future
As AI matures, every industry will consolidate around a few deep, data-rich ecosystems—not a thousand disconnected tools. The great rebundling is already underway. Healthcare, logistics, law, energy, construction—each is finding its own AI-native stack.
This is the quiet revolution of 2025: SaaS is no longer software-as-a-service.
It’s systems-as-intelligence.
