NextFin News - Orlando Bravo is trying to reset the logic of software buyouts in the middle of the AI transition. The founder and managing partner of Thoma Bravo is arguing that AI is not simply a threat to enterprise software economics, but a new layer of revenue that the best software owners can package, govern and sell. That matters because Thoma Bravo is not making a symbolic bet. The firm manages almost $200 billion in assets, and its portfolio companies generate around $3 billion in annual revenue, giving Bravo a large enough platform to test whether AI is a tailwind for mature software or a force that strips away the old rent model.
The core claim is striking in its simplicity. Bravo says roughly 50% of Thoma Bravo’s new revenue comes from AI and agentic revenue. He also says the “SaaSpocalypse” is over and that AI is an “enormous tailwind” for software companies. Those lines are doing a lot of work. They are not just defending software valuations. They are trying to convince investors that the enterprise software stack can absorb AI before AI turns the stack into a commodity.
That argument emerged after a year in which software investors became much more sensitive to the possibility that AI could bypass traditional interfaces, automate routine tasks and weaken the pricing power of subscription products. The fear was not that software would vanish. It was that the easiest layer of software to copy, automate or bundle would become less valuable, while only the deepest workflow platforms would keep their economics. Bravo’s answer is that software companies with real distribution, data and workflow ownership can make AI part of the product rather than a substitute for it.
That is why the story matters beyond one executive’s comments. If Bravo is right, then AI does not destroy software buyout economics. It changes the mix of what private equity has to underwrite: less pure subscription durability, more product reinvention, more data governance and more proof that automation creates cash flow rather than just a demo. If he is wrong, the buyout model faces a harder test, because the value that once came from recurring software seats may migrate to a different part of the stack.
What The “SaaSpocalypse” Debate Is Really About
The market argument around software has been framed too crudely as a fight between believers and skeptics. In practice, it is a question about where value sits inside a workflow. Software companies that sit close to the customer’s actual business process can use AI to broaden the task set they cover. They can turn copilots into agents, agents into workflow orchestration, and orchestration into a stronger reason to pay. In that case, AI expands the product surface and improves retention.
But not every software layer has that kind of power. If a product mainly offers a thin interface over generic functions, AI can compress its economics. Customers may still need a tool, but they may not need as many seats, modules or implementation hours. That is the risk behind Bravo’s rejection of the “SaaSpocalypse” idea. He is implicitly drawing a line between software that owns the workflow and software that merely presents it.
The reason his view carries weight is that Thoma Bravo lives inside this distinction every day. The firm’s long-running buyout strategy has been to acquire mission-critical software businesses, improve the operating model, push recurring revenue and exit at a higher multiple. That formula worked because enterprise software was sticky and slow-changing. AI shortens the product cycle, raises customer expectations and changes the economics of feature delivery. The winner is no longer just the owner of recurring revenue. It is the owner of the data, the workflow and the permissions layer that makes automation useful.
Bravo has said the firm’s portfolio is already seeing that shift in revenue terms. If around half of new revenue is AI and agentic revenue, then AI is not a future promise inside the portfolio. It is already getting monetized. That does not mean the economics are settled. It means the private equity case is now being tested in real time.
“Around 50% of our new revenue is AI revenue, agentic revenue,” Orlando Bravo said.
That sentence is important because it turns AI from a narrative into a line item. Investors can debate the durability of the number, but they cannot dismiss the fact that a major software investor is already attributing a meaningful share of new sales to AI-linked products. That suggests the question has shifted from whether AI matters to software to which software owners can turn AI into repeatable revenue before the market starts pricing the category differently.
Why Private Equity Has To Rethink Software Underwriting
The buyout model for software has long relied on a familiar set of assumptions: recurring revenue, mission-critical status, low churn and steady expansion. AI does not erase those assumptions, but it makes them less sufficient. Private equity owners now need to understand model strategy, data access, security, compliance, integration cost and the pace of product change. Those are not side issues. They determine whether a software company can defend pricing and justify the multiple it needs at exit.
For Thoma Bravo, that means AI is both an opportunity and an operational burden. A firm that specializes in buying and improving software has to ask whether AI can be layered onto existing products fast enough to preserve the moat. At the same time, the firm has to budget for the fact that AI can raise the cost of product development, increase the need for governance and expand cybersecurity exposure. Bravo has acknowledged those questions publicly, warning that governance, cybersecurity and returns remain open issues even as he argues that AI will help software.
That is the more nuanced reading of his stance. He is not saying every software company wins. He is saying the companies with genuine workflow control can convert AI into a revenue engine, while the weaker ones may see their economics eroded. That distinction is becoming central to buyout underwriting because AI can change how an application is sold, how it is priced and how it is used inside the customer’s business.
The labor angle matters too. Bravo has also argued that AI can make junior workers’ jobs “greater” and “broader” by automating grunt work. In software terms, that hints at a future in which routine tasks move into automation while humans move up the value chain. For vendors, that can mean faster implementations and higher-value services. For investors, it means the margin story is no longer just about cutting costs. It is about redesigning the organization around a faster product cycle.
The private equity implication is straightforward: AI is no longer a layer you can ignore in diligence. It has become part of the core underwriting question. If a sponsor believes the product can become more valuable because AI expands the workflow, it may pay a higher multiple. If it believes the product can be substituted by an agent, it may pay less. That is why Bravo’s framing is more than optimistic marketing. It is an attempt to redraw the valuation map for an entire corner of the market.
Why The Market Still Treats Software As A Battleground
Public-market investors are still trying to sort software into winners and losers, and that is exactly why the category remains volatile. Some companies will use AI to raise engagement, improve retention and deepen their share of the customer workflow. Others may see the opposite: lower seat counts, more pricing pressure and greater customer skepticism about what the software layer is actually worth.
The key difference is whether the company owns the part of the stack where the work gets done. If it does, AI can become an upgrade cycle. If it does not, AI can become a substitute. Bravo’s thesis is that software companies with embedded data, customer trust and workflow control can sit on the right side of that divide. That is a defensible position, but it is narrower than the broad claim that “AI is good for software.” It says the value will be redistributed, not universally preserved.
That redistribution matters for private equity because exit timing, product investment and pricing assumptions all depend on it. A sponsor that thinks AI will make a portfolio company more strategic may keep investing and wait for a better sale process. A sponsor that thinks AI will compress the product moat may move faster to exit. In both cases, AI is now part of the capital-allocation decision, not just the product roadmap.
The market has already shown how quickly sentiment can turn when a new AI tool appears capable of doing work once associated with software. Investors have been willing to mark down some software names when they fear an interface can be bypassed by an agent. But they have also been willing to re-rate companies that can prove AI improves usage, broadens the task set or makes the platform more indispensable. The challenge is that the market does not reward “AI exposure” in the abstract. It rewards evidence that AI creates measurable economic value.
That is what makes Bravo’s comments strategically useful. They push the discussion away from a binary view of disruption and toward a more precise model of where the money actually sits. AI may compress the weakest layers of the stack, but it can also deepen the economics of the strongest ones. The point of the reboot is to identify which is which before the market does it for them.
What Comes Next
The next test is whether Thoma Bravo’s portfolio can keep turning AI features into revenue without sacrificing margin discipline. Investors will want to see whether customers pay for agentic capabilities, whether renewal behavior improves or deteriorates and whether software owners can keep the economics intact as product teams spend more on automation, security and data infrastructure. Those are the metrics that will determine whether Bravo’s argument is a temporary defense or a durable shift in the software-buyout model.
The broader market will also watch whether the AI revenue cited by Bravo remains concentrated in a handful of products or broadens across the portfolio. If it broadens, software buyouts may look less vulnerable than skeptics feared. If it remains narrow, the thesis will be seen as more selective, tied to a few winners rather than the entire category.
For now, Bravo is making a clear bet: the companies that own the workflow can keep the economics, while the ones that only rent the interface may not. That is not a slogan. It is a map for how software investing may work in the AI era.
Explore more exclusive insights at nextfin.ai.