I’m a CEO who oversees .5 trillion in spend data. AI’s winners are already decided

By Leagh Turner, CEO of Coupa

The Debate on AI and Enterprise Software

The ongoing discussion regarding the impact of artificial intelligence (AI) on enterprise software often misses the core issue. Prominent leaders from companies such as Intuit, Salesforce, and Box have defended the role of Software as a Service (SaaS) in the evolving landscape of AI. Thoma Bravo’s Holden Spaht recently remarked, “Software is AI if you do it right.” I would take this a step further: SaaS and AI are not distinct entities; they are both facets of software. Some SaaS companies, particularly those with strong historical foundations, can achieve significant synergy by rapidly leveraging AI technologies.

The Role of Data in AI

To understand the dynamics at play, consider the analogy of two architects. One has read every book on structural engineering, while the other possesses those same books along with blueprints, soil samples, and maintenance records for every building in a specific area over the last two decades. Who would you trust to construct a skyscraper in a seismic zone?

In this analogy, AI serves as the engine, and data is its fuel. However, not all data is created equal. Software that relies on generic, unstructured, and unverified public internet data is akin to using low-grade kerosene to fuel an engine. In contrast, software platforms designed for specific purposes, which manage critical workflows and have accumulated trusted, proprietary data over time, operate on high-octane fuel.

At Coupa, this translates to overseeing $9.5 trillion in proprietary transaction data generated by over 10 million buyers and suppliers engaged in real business transactions in real time.

The Unfair Advantage of Early Investment

With over 20 years of experience in enterprise technology, including tenures at Xerox, Oracle, SAP, and Ceridian, I can confidently assert that the companies currently succeeding in the AI landscape did not begin their journey with the advent of ChatGPT three and a half years ago. Instead, they laid the groundwork a decade ago through investments in machine learning and predictive analytics.

For years, we have been “priming the pump,” utilizing machine learning to clean data, categorize spending, and identify risks. This foundational work creates a significant difference between being a “bolt-on” solution and a “built-in” one. Software vendors who are only now discovering AI are essentially attempting to attach a jet engine to a horse-drawn carriage, lacking the structural integrity required for success.

The Divergence in the Market

As we look to the future, a clear divergence is emerging in the market:

• The Losers: SaaS providers that merely serve as thin user interface wrappers over public models. They lack the deeply embedded workflows necessary to deliver real customer value and will struggle to evolve quickly enough to remain competitive.
• The Winners: SaaS providers with domain expertise, deeply integrated workflows, and the capability to autonomously manage critical actions—such as tax compliance, supply chain resilience, and fraud detection—while also assisting customers in transforming their workforce.

It is important to note that great technology that users resist will simply become shelfware, failing to deliver the intended benefits.

The Shift Towards Outcome-Based Pricing

We are witnessing a fundamental shift in how software is purchased and sold. The traditional concept of the “seat license” is becoming outdated. Why should a company pay for a password when the focus should be on paying for results?

Progressive leaders are already adopting pricing models centered on outcomes. For instance, 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 rather than simply the number of employees with access.

The Future of Software and AI

The timing of this transformation—both in technology and workforce—remains uncertain, but the pace of change is rapid. Software vendors that have only recently embarked on their AI transitions may find themselves at a disadvantage. For organizations aiming for success, innovation must begin with data. The future will not belong to the most advanced AI model; it will belong to the software that possesses the best data and is deeply rooted in the daily operations of companies.

Conclusion

The landscape of enterprise software and AI is evolving rapidly, and the companies that have prepared themselves through strategic investments in data and technology are poised to lead. As the market continues to differentiate between those with a robust data foundation and those without, the importance of leveraging AI effectively cannot be overstated.

Note: The opinions expressed in this article are solely those of the author and do not necessarily reflect the views of Fortune.