It’s still early days for AI adoption, but one thing is already evident: While AI activity is everywhere in private equity, meaningful value creation isn’t. Let’s call it the AI Value Paradox.
This isn’t unique to private equity, either: Bain & Company’s most recent CEO survey finds that roughly 80% of chief executives across the economy are unhappy with the pace of their AI transformation programs despite frenetic activity.
It’s true that software companies and software-intensive businesses are seeing more traction because AI can change the economics of software engineering itself. But for most portfolio companies, there’s little correlation between AI spend and value, despite the very real promise offered by these transformational technologies.
Our conversations with PE deal teams and portfolio company leaders reveal several recurring issues that are preventing progress.
Use-case swirl. Organizations today are generating hundreds of theoretical ideas for deploying AI, with new ones appearing every week. But few have disciplined ways to prioritize them, and spreading resources across a random collection of experiments means none of them get enough attention to move the P&L.
The micro-productivity trap. Haphazard AI adoption may save employees a few minutes drafting emails, summarizing documents, or completing administrative tasks. But while the company becomes incrementally more efficient, enterprise value doesn’t budge.
Tools-first bias. Teams begin by selecting an off-the-shelf copilot or responding to a vendor pitch. They layer the tool onto an existing workflow instead of redesigning the process end-to-end with AI capability in mind. Build-vs.-buy choices are inconsistent. Fragments get automated without a meaningful boost in productivity.
Capability gaps. The organization lacks the data, talent, or technology to move from prototype to production. It launches broad-brush capability programs that consume time and capital without proving value. Momentum fades before results can land.
Sponsorship gaps. AI is delegated to the technology organization, innovation team, or functional pilots. Without visible ownership at the top and explicit links to the company’s value-creation plan (VCP), conviction never forms and the work stalls—not because anyone rejected it, but because no one was convinced it would last.
A path to real value
The common denominator here is a disconnect between what the portfolio company aims to accomplish strategically and a practical assessment of how AI can help. The companies getting AI right aren’t investing behind the shiniest list of use cases. Instead, they start with the VCP and determine how (if at all) AI can be transformational in driving its most vital initiatives. This leads to a focused, CEO-level agenda that the company can rally behind to steer execution, catalyze change, and make the new ways of working stick.
Consider how this worked at a PE-owned company that specialized in managing large properties in the hospitality and leisure industry. It had grown rapidly but needed to evolve its business processes to maintain its momentum.
While AI often lends itself to cost-cutting, the company’s VCP revolved around revenue: retaining existing clients and adding new ones by stepping up its service game. So, while management had identified 30 possible processes that might benefit from AI intervention, it quickly narrowed the list to 4 linked closely to revenue improvement: finance automation, lead generation, proposal automation, and client value reporting.
The CEO championed an effort to break down the relevant workflows and determine specifically how AI could transform them. Automating finance and accounting might not seem like an obvious revenue lever, but in this case it was all about retention, enabling the company to provide better financial information faster to its clients, which was a mission-critical element of its value proposition. AI also allowed it to systematize production of bespoke reports detailing how much value the company was bringing to these complex properties, greatly enhancing marketing communications.
In both use cases, AI turned a manual, highly fragmented process into an automated, scalable workflow that could draw on approved data, format it appropriately, and conduct rules-based validation in a fraction of the time it took previously. Now the company is tracking to deliver significant near-term return en route to a 40%–50% boost in EBITDA by 2030.
Dialing in a focused plan
This company broke free of the AI swirl that ensnares so many others by focusing on four key steps.
Start with the VCP and hunt for home runs. It’s easy to get distracted by what other companies are doing with AI, but all that matters is how your organization can deploy AI to solve your biggest issues. Isolate the three to five key opportunities embedded in the VCP and carefully map those workflows to determine how AI can transform them into something better.
Reinvent the workflow before you pick a tool. This involves reimagining the future-state version of your priority workflows as if you were building them from scratch with AI already in hand. Which activities should disappear? Which activities can be automated or augmented, and where do people still add distinctive value? Name specific enablers or barriers to adoption—things like data readiness, technology architecture, talent issues, or operating model limitations.
Blueprint and sequence the build-out. Next comes translating each redesigned workflow into concrete product requirements: What does the solution have to do, what data does it depend on, and what quality bar must it clear? Prototypes and rapid iteration are essential here. But once they’re validated, concentrate resources on moving them to production, not parking them as pilots.
Manage the change. Achieving scale involves robust product development and change management. For one large parts distributor, AI radically shortened the average turnaround time for order inquiries from more than an hour to less than five minutes. But the solution disrupted the status quo. The company chose to build an AI-enabled tool that replaces manual research by ingesting reams of unstructured information from parts catalogs, equipment manuals, and OEM data to produce a unified, searchable knowledge base, giving customers faster, more accurate parts recommendations.
Getting there involved four months of initial use-case assessment and an all-hands-on-deck production effort endorsed by the CEO. While the investment was significant, the payoff is showing promise: Although still early, the company is seeing its conversion rate double, and the function is scaling nicely without adding to headcount.
Momentum builds on itself
Once the organization anchors around a few CEO-level priorities, capital follows conviction, and everything else compounds from there. Early results lend credibility, momentum builds, and focused investment translates into earnings—exactly what today’s deal math demands. A lot of ideas are great. But a single home run use case that reaches the P&L creates more value than dozens of pilots ever could.