“IT and the business are one” is the craziest great idea in enterprise tech
From August 2025
Are IT and business one? This is a recurring idea on enterprise technology management. Impressionistically, we’re probably in an “upswing” era for this idea than a “downswing era” for this meme -- but is it true.
1. At one level, it’s not only true but obvious. Technology is an essential business function, no less a part of the business than strategy, marketing, sales, operations, etc.
2. The language around “IT” versus the “business” is neither helpful nor unique among business functions. Many times I’ve sat in a meeting a CIO and his or her “HR partner” and heard the person from HR refer to IT as “the business.” Corporate strategy teams talk about their interactions with “the business” all the time. Language matters, but only so much. Sometimes “the business” just provides a convenient linguistic shortcut to the idea that a business function has internal customers it must support.
3. When we talk about “IT as part of the business” we conflate two issues technology as an organization and technology as a discipline.
4. Different businesses will always choose different constructs for where to house technology, including how centralized or centralized technology personnel will be. But -- saying “IT should just be in the business” is a counter-productive, and almost meaningless statement. Put all of technology into the business units means forgoing any scale in infrastructure and cybersecurity or any value from, for example, shared data. What’s more, what does “in the business mean?” For a global bank, does “in the business” mean in the retail business unit, in the consumer line of business or with the mortgage product?
5. In contrast, technology engineering and operations is a discipline (or a collection of disciplines). Building and running systems that are functional, performant, scalable, efficient, secure, compliant and resilient is a skill honed over decades. No amount of vibe coding will help a business manager do this.
6. But what about determining what to build? Not too long ago the answer to this question was easy. Use a product operating model to separate accountability for what to build and how to build it. Business leaders defined the (business-facing) portfolio and created requirements. Technology leaders (in a well-designed model) had the authority and autonomy to execute to specifications. And this worked pretty for automating order-to-cash processes, for example.
7. But now we face a massive disruption of how enterprises might use technology, given GenAI (plus agentic workflows, plus knowledge graphs). For the first time in decades cutting edge computer science is reshaping how enterprises might get value from technology. A CIO or CTO has a hell of a lot better chance of understanding a sales process than a sales VP has of understanding how likely evolution of frontier models might create opportunities for sales coaching. (Obviously, the dynamics are a bit different for pharma manufacturing or high frequency trading, where the business domain knowledge is infinitely more complicated.)
8. HFT may be instructive -- in some ways “the business” and “technology” have become one. We’ve all sat in meetings with people started out engineering trading algorithms now creating trading strategies. Within these companies, systems engineering remains a distinct discipline, business and technology have come together, not because IT has been subordinated to the business, but because technologists have assumed responsibility for “the business.
”