Executive Summary
The next software exit will be priced on whether the product earns the customer's trust to do the work. It will not be priced on another point of margin. Margin and growth discipline remain necessary, but in the agentic era they no longer tell a buyer the asset will last.
For 15 to 20 years, the PE software playbook has been reliable. Buy recurring revenue, professionalize sales, raise price, take out cost, add acquisitions, and sell on EBITDA and the Rule of 40. That playbook treated product strategy as an expense line and a roadmap slide. It worked because switching costs protected the revenue while the product stood still.
That protection is weakening. PE and strategic buyers now ask a different first question: will this product still matter in 3-5 years when frontier labs build competitive products, AI first competitors suck up the oxygen in the room and/or DIY IT initiatives powered by LLMs consume the available budget?
I have spent the last 15 years as a Chief Product Officer most of it answering to PE boards, through numerous exits. My argument has two parts.
Stop bolting agents onto legacy products. A chatbot beside an unchanged workflow is table stakes. Value comes from agents that understand the domain, the company and the user, propose the work, and earn the right to do it, until human review is the exception and not the rule.
Pay down the product debt the old playbook deferred. Feature factories, top-down mandates, skipped design, slow time to value, unfunded architecture and unintegrated acquisitions all compound. Each one is now visible in diligence.
Neither part requires a rebuild or a bigger R&D budget. Both require a different operating model: small teams of domain experts, disciplined discovery, a governed semantic foundation, and a board that manages product strategy as its largest investment. A long hold is usually discussed as a problem. I think it is an opportunity to do this work.
