How enterprise software companies survive the AI apocalypse

Mar 3, 2026
AI

Enterprise software companies that provide unique data, shared responsibility, scale, and stand above the token & time tradeoff, will thrive in the evolved marketplace.

We are going to see industry consolidation.

The enterprise software vendors that will grow and acquire competitors during this period of consolidation, will:

1) Produce unique, multi-customer data, and manage it correctly

Multi-customer data:

Software vendors need to process more unique data than the customer could capture from their own environment alone. Enabling the vendor to predict trends that individual customers could not anticipate with their siloed dataset.
This requires massive telemetry across wide customer installation bases.

Proprietary data:

Proprietary, primary source research, produced exclusively by the vendor.
Content that the vendor creates or has exclusive licensing for.

Correctly managed data:

Confidentiality, Integrity, Availability - A vendor with a superior dataset must have the technical capital, monetary capital, and reputation to ensure (and instill confidence) that the dataset will remain available, unaltered, and confidential to the customer.
Data sovereignty
- In a time of global political tensions, solutions that can be scaled across multiple private data centers will prevail.

Unique datasets will be the largest durable differentiator for enterprise software companies. Even further, enterprise software has a bright future in the world of AI, because customers will continue to be looking for partners that additionally:

Support, SLAs, professional services, and cyber insurance cost reductions.
Organizations that choose to build critical technology solutions independently also assume full liability for the continuous, effective operation and maintenance of that solution.
When something goes wrong with a mission-critical system that a customer has built themselves, who can they call?

A note on professional services: Now is prime time to sell professional services, as organizations look for assistance adopting and implementing AI into their ecosystems. The window to sell professional services is closing. On the long horizon, services will be heavily replaced by AI agents. However, the window is closing slowly, services will provide a healthy source of revenue for vendors for at least the next half a decade.

3) Offer solutions that scale

While software built with AI code generation tools can be developed quickly for proof of concepts, these solutions do not scale or have the long tenure in operational use to demonstrate they can meet the demands of the modern enterprise.

4) Stand above the token & time tradeoff (t&t/t)

In the previous decade, we have seen many organizations transition away from building internal systems themselves and instead buy commercially available products. The AI revolution will move the slider on the spectrum more towards self-built solutions for low-risk internal use cases.

However, every determination to build something internally will be met by two immovable constraints. One present since the dawn of human labor (time) and a new monetary construct introduced by AI coding engines (tokens).

For a nontechnical example of the t&t/t, consider: While many homes can acquire the technology necessary to produce beer domestically. In practice, very few do. Doing so requires prioritizing time and prioritizing resources. Most consumers opt for prepackaged beer because they can acquire it quickly and be assured they receive a complete, high-quality product immediately and reliably.

The same is true with software. While some organizations may hold the view that they can build solutions internally;

Will they have the time to do so?
How will they determine which projects get funded with the tokens necessary for them to be delivered?
What trade-off decisions are they making by choosing to spend their time building these capabilities themselves?
What projects core to their business model are they distracted from by investing time and resources into the process of trying to learn from scratch how to arrive at the completed product?

Organizations may find that building their own capabilities to ferry insights between their unique business structures and within their unique business rules may be worthwhile. They will, however, find the massive investment necessary to build foundational architecture that must operate at scale and with extreme reliability not worth the trade-off.