The reason comes down to one thing: data.

The Architect and the Fuel

Imagine two architects. One has read every book ever written on structural engineering. The other has those same books — plus the blueprints, soil samples, and maintenance records for every building in a specific area for the last twenty years. Who do you trust to build a skyscraper on a fault line?

AI is an engine, and data is its fuel — but not all fuel is equal. Software built on generic, unstructured, unverified public internet data is essentially feeding AI low-grade kerosene. Software platforms that serve a specific purpose, handle mission-critical workflows, and have amassed trusted, proprietary data over long periods of time are running on the highest-octane fuel available. At Coupa, that means $9.5 trillion in proprietary transaction data — generated by more than 10 million buyers and suppliers conducting real business, in real time.

The Unfair Advantage

I’ve spent more than 20 years in enterprise technology — at Xerox, Oracle, SAP, and Ceridian before joining Coupa — and the companies that are winning today didn’t start their AI journey with the launch of ChatGPT three-and-a-half years ago. We started a decade ago, laying the foundation for this critical moment with machine learning and predictive analytics. For years, we’ve been “priming the pump”—using ML to clean data, categorize spend, and flag risk. The result is the difference between being a “bolt-on” and a “built-in.”

If a software vendor is just now discovering AI, they are essentially trying to install a jet engine on a horse-drawn carriage. The structural integrity isn’t there. The companies that will win are those that have been building the data foundation for this moment and are moving fast to incorporate AI into the core of the product.

There will be a clear divergence in the market:

• The Losers: SaaS providers that function as thin UI wrappers over public models. They lack the deeply embedded workflows necessary to deliver tangible customer value, and they won’t evolve fast enough to compete.
• The Winners: SaaS providers with domain expertise, deeply embedded workflows, and the ability to autonomously manage mission-critical actions — tax compliance, supply chain resiliency, fraud detection — while also supporting customers in transforming their workforce. Great technology that people resist using simply becomes shelfware.

The Move to Outcomes

We are seeing a fundamental shift in how software is bought and sold. The “seat license” is a dated concept. Why should a company pay for a password when they should be paying for a result?

Forward-thinking leaders are already adopting pricing models focused on outcomes. If our AI identifies $10 million in duplicate invoicing — and our customers have already realized over $300 billion in cumulative lifetime savings on the Coupa platform — our revenue should reflect that realized value, not just how many employees have a password.

The timing of this total transformation — technology and workforce alike — remains uncertain, but the velocity is staggering. Software vendors that only recently started their AI transitions may find themselves at a disadvantage. For organizations that plan to win, innovation has to start with data. The future doesn’t belong to the smartest model — it belongs to the software that has the best data and the deepest roots in a company’s daily operations.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.