AI-Native, Not AI-Sprinkle: Why AI Is A Business Change, Not A Technology Change

The buy-and-build SaaS playbook regularly faces the problem of old code: A roll-up strategy executed over time accumulates separate aging code bases from the acquired businesses.

A clean sheet rewrite of a legacy product certainly improves customer experience, but it can take years from starting gun until the last customer is migrated and the old code is fully decommissioned.

is an HR software business, owned by my investment firm, with just that challenge. In December, when joined as CEO, the company had developed a plan to rewrite from scratch one of its oldest software products. The timeline was 18 months, with a 30% surge in engineering headcount to power through the project. But Jeff and his new CTO did it better, faster and smarter.

Jeff came to the board in February with a radical alternative: redesigning the engineering team organization and individual job specs, literally changing what people do all day to best put to work the power of off-the-shelf AI tooling.

HireRoad鈥檚 new approach would complete the development in 16 weeks, not 18 months, and the customer base migration and legacy decommission would be completed in calendar year 2026. In my 30-year career as a software investor, I had never seen any organization achieve such a task at anything like that velocity. The board debated and made the leap, killing the old plan and taking this frontier bet.

The rebuild was done in 15 weeks, a week ahead of schedule, and as of this writing, the first 34 customers have been migrated to and are live on the new platform, with glowing feedback. The pacing to complete the migrations and decommissioning is on track. The kicker is that Jeff and team completed this with a smaller team, freeing up the 30% headcount surge to work on other HireRoad developments.

AI-Sprinkle vs. AI-native

In 2024 and 2025, we at felt proud of ourselves and quite cutting-edge for providing the engineering teams across our software portfolio with access to AI tools such as Copilot and Claude Code. We saw productivity gains of 10%, then 20%, now more like 30%.

But somehow, our companies were all getting stuck at those 30ish percent gains.

How to reach 3x gains? The realization was that providing AI tool access alone was, candidly, not AI-enabled but rather AI-sprinkled. The breakthrough came when leaders went beyond the AI-sprinkle and instead adopted AI-native daily practices.

Let鈥檚 pause for a moment on terminology here. The phrase 鈥淎I-native鈥� is thrown around a lot just now. In our usage, AI-native describes what you do all day, not when your company was founded. Anyone can learn to work in an AI-native fashion, and it means directionally using AI tooling first and humans to orchestrate, coordinate and communicate.

AI-native work is not just doing the same thing faster; it means doing different things with more delegation and quicker learning loops, and I will share some specific examples as we go.

Startups will call the move to so-called AI-native organizational practices obvious. They are right, but they are not burdened by an existing organization or established products and customer bases. They get to build AI-native practices into their organization from the start. In contrast, private equity portfolio companies have to remodel.

Our experience is that the AI-sprinkle 鈥� or, giving an AI layer to an otherwise unchanged organization 鈥� provides mere percentage gains to productivity. We have to redesign the organization around the power of the tools to get multiples on productivity.

A 30% productivity gain feels good, but it is the trap of the current moment in AI. And the path from 30% to 3x is uncomfortable. It runs through changing how teams are structured and what people actually do all day. In this way, delivering on the promise of AI is a business change, not a technology change.

I had the great good fortune to take a course in strategy at business school from and Andy Grove. Burgelman is a professor whose 12-year study, , delivered the definitive business text on , which Grove famously ran through its own era of technological revolution in the chip industry. His intellectual framework applies exactly to the current moment of technological revolution.

Evolutionary vs. revolutionary

Burgelman鈥檚 framework is that there are two kinds of strategic behavior, which he called induced and autonomous. Induced strategies fit the company鈥檚 existing structure and trajectory, like an AI layer inserted into an existing process. They are evolutionary moves, continuously advancing and improving on the current direction of travel. Autonomous strategies are those arising from outside the current business plan, like rewriting the job definitions and changing the team structure and work patterns of your product and engineering teams around the power of AI tooling.

Autonomous strategies are revolutionary moves. With AI, 30% gains are to be had from AI-sprinkle on the induced-strategy evolutionary path. The 3x gains require AI-native autonomous strategies, meaning revolution.

An oft-repeated analogy is how electricity transformed manufacturing. Replacing the steam engine powering a mill with an electrical motor delivered very little productivity gain.

Productivity skyrocketed only when the manufacturing plant itself was redesigned, distributing small electric motors throughout the factory in a horizontal layout, delivering what a single steam engine never could. What interests me most about this story is why it took decades before the factories were redesigned. Why couldn鈥檛 those organizations make the revolutionary leap more quickly? That is where the Burgelman/Grove case study is so helpful.

Burgelman points out that revolutionary ideas are very often squelched by institutional inertia and the cultural power of the evolutionary path. To be realized, revolutionary strategies need full buy-in from the CEO and Board.

The retelling of Grove鈥檚 revolutionary moment is here very apt. As told in Grove鈥檚 seminal business book 鈥�,鈥� he and Intel co-founder were sitting together struggling with a strategic question. Intel鈥檚 primary business at that time was memory chips, a business where Japanese competitors were assaulting them in a brutal price war, pushing Intel to the brink. Intel also had a smaller, growing business line in microprocessors, the CPUs inside personal computers.

After a long pause, head in hand I imagine, Grove looked up at Moore and said, “If we got kicked out and the board brought in a new CEO, what do you think he would do?” And Moore said without hesitation, 鈥淗e would get us out of memories.鈥� Grove replied, in effect, why shouldn鈥檛 you and I take a walk around the building just now, and come back in the door, and do it ourselves?

That is just what they did, and the great run of 鈥淚ntel Inside鈥� as the leading CPU maker was launched. The uprooting of your proven daily practices and time-tested organizational design, to an AI-native way of working and team design, is a difficult revolutionary act. It may feel just as uncomfortable, just as heroic, as that fateful Grove-Moore conversation.

So, what did HireRoad do to affect the 30% to 3x revolution? The new technology leadership trained the team on a new hour-by-hour how to spend your day, built around the power of the AI tooling. The new sales leadership worked with the engineers to put the rapidly produced prototypes in the hands of clients, shortening the user feedback loop. When users identified bugs, the system logged them, wrote code to fix them, and presented the solution to a 鈥渉uman in the loop鈥� for final judgment and publication. Customer support was engaged to develop and communicate a high confidence transition plan for users.

Overall, the HireRoad team became smaller and more senior, with resources freed to work on other initiatives, and to roll out these practices across other HireRoad product lines.

Management innovation and private equity

AI-native organizations are the third major management innovation of my private equity career. The first management innovation was the removal of bloated cost structures and tight linkage of executive compensation to equity outcomes in the 1980s, and the second was the conversion of on-premise licensed software to subscription model SaaS in the 2010s.

Those investors who mastered and first put those techniques into practice created vast fortunes for their capital partners. The starting gun has just been fired on the third wave. The organization changes to implement AI are a business change, not a technology change. While the ideas and practices can arise from anywhere in the organization, companies will not participate until this revolutionary change is endorsed by the CEO and board.

There are some 10,000 privately held software companies in the U.S. today, depending on exactly how you count. Leaders of those businesses know, explicitly or perhaps just through gut feel of the shifting sands, that doing the same thing in the same way in the age of AI is a losing strategy. You won鈥檛 lose all at once. You will be slowly starved as competitors move at 3x your pace around you. Certainly, your prospects to be a leader will close.

You have the customers, the distribution and the knowledge of the problem you are solving, all legs up on the startups. The nature of the organizational change you need to make is known, or knowable.

When considering this moment, shared by all of us who work with existing software organizations, think about the decades between the initial one-big-motor electrification of factories and the 1920s many-small-motors factory redesign which delivered the huge productivity gains. These changes don鈥檛 just happen on their own, and this time around, we won鈥檛 have the luxury of a lengthy transition. When considering your own revolutionary strategic move, run the Grove thought experiment. Walking outside around your building, ask yourself, 鈥淚f I were fired, what moves would the newly hired CEO make today, to win with this company in the age of AI?鈥� I suspect the nature of your answer will not be to sprinkle more LLM access across your unchanged organization. Rather, ideas will occur to you on how to change your team structures and what people do all day to better serve your customers through the incredible AI tooling now at your disposal.

Are those the moves you are making today?

co-founded in 2014 and is managing partner. He has served on numerous private and public technology company boards, and currently is a director of , , , , and . Previously, he was a partner and member of the investment committee at . He also worked at and . Morse serves on the board of directors of and as member of the advisory board for the HMTF Center for Private Equity Finance at . He attended , graduating summa cum laude with a BSE, and , where he earned his MBA and was an Arjay Miller Scholar. Morse lives in Austin.

兔子先生传媒