In year two of generative AI’s sprint across the technology landscape, it’s clear everybody in private equity is thinking hard about putting this innovation to work. Today’s winners are figuring out where AI can deliver meaningful results and how to build organizational support for AI adoption.

When we surveyed private investors representing $3.2 trillion in assets under management in September 2024, they reported that a majority of their portfolio companies were in some phase of generative AI testing and development, and nearly 20% of companies have operationalized generative AI use cases and are seeing concrete results. This is an impressive result for such a nascent technology.  

It’s not surprising that the preponderance of use cases remain in pilot mode—most firms report that companies are still getting up to speed with the plethora of new tools. But that’s as it should be. Today’s generative AI models are not well suited to every task, and firms need to figure out the right use cases through experimentation. 

The firms having the most success mining value tend to share a similar outlook: They’ve become true believers in generative AI’s potential and are committed to managing decisively through this period of change and ambiguity.

These firms recognize that they don’t—and can’t—know everything yet about how generative AI might deliver value across their portfolios. But they’re not letting that get in the way of taking concrete action. They are investing aggressively to build firmwide expertise and helping their portfolio companies apply the power of AI to their most important strategic initiatives. And, together with their management teams, they are tackling change management challenges to overcome “organ rejection” among employees resistant to technologies that could threaten their jobs.

Working from the principle that what’s not possible today may be game-changing in very short order, PE firms are organizing to innovate in several important ways:

• At the fund level, they are rapidly assessing the current strengths and weaknesses of generative AI technology while organizing themselves to learn systematically and share those insights with and between portfolio companies.
• They are investing in tech capabilities, adding AI talent, setting up governance protocols, and assembling experts and advisers to help both the firm and portfolio companies stay attuned to what’s on the horizon.
• At the same time, to avoid unfocused dabbling, they are challenging portfolio company management teams to identify a short list of top business priorities and tapping every available resource to determine how AI could help push those initiatives forward. Importantly, they view AI as a tool in service of strategy, not a strategy on its own.
• Their portfolio companies are then scoring early return on investment by using AI to enhance products, boost revenue, and expand margins via operational efficiencies. They are making fast, pragmatic decisions on whether to build, buy, or partner to develop solutions, and they’re moving forward buttressed by clear strategies for implementation and change management.

Learning and doing at the same time is never easy. But here are some field-level glimpses of how several firms are approaching the challenge.

Bringing the full-court press

It’s safe to say that software specialist Vista Equity Partners has gone all in on generative AI. Firm leaders are already convinced AI represents a paradigm shift in innovation that will ultimately create a multitrillion-dollar investment opportunity. In fact, over the next three to five years, Vista expects that AI’s outsize impact on a software company’s top and bottom lines will rewrite the Rule of 40—the yardstick investors have used for years to evaluate promising software-as-a-service (SaaS) companies. As AI helps the industry enhance products and cut costs, Vista believes the new standard for revenue growth plus margin will reach 50% or even 60%.

Armed with that conviction, Vista is marching into battle. To supercharge learning and implementation, the firm has arrayed an internal army of professionals dedicated to helping its 85-plus portfolio companies apply AI across the organization in product innovation, research and development, go-to-market, talent, and operations.

As part of its annual operational planning process, Vista is requiring each of its portfolio companies to submit goals and quantified benefits from generative AI initiatives. It regularly screens and triages its portfolio to determine where opportunities and risks lie and then partners with management teams to help them either move out of harm’s way or seize on the potential to enhance value. Company executives are expected to share what they’re learning at a GenAI CEO Council organized so small companies can learn from large ones and vice versa. Vista’s team of experts then works hand in hand with management teams to accelerate adoption and results, defining and monitoring relevant metrics along the way.

Vista has even found a way to “gamify” AI adoption in its portfolio through hackathons, held annually in the US and India, where companies compete to develop the best use cases using large-scale, pretrained generative AI tools accessed via industry partnerships. Hackathon projects launched just under two years ago have already become revenue-generating products at scale today. 

As a software specialist, Vista has a clear need for generative AI at nearly all of its majority-owned portfolio companies—most, at a minimum, are using AI-based code-generation tools (driving up to 30% increases in coding productivity for scaled adopters). But 80% of those companies are also deploying generative AI tools internally or developing new products. A good example is Avalara, a Vista portfolio company that makes tax compliance software. It is using a generative AI tool from Drift (now part of Salesloft) to increase sales rep response time by 65%.

Another case study: LogicMonitor, which offers AI-powered data center transformation software. The company’s SaaS-based monitoring platform uses generative AI to summarize complex alerts from multiple sources of data either on premises or in multi-cloud environments. It pinpoints existing problems rapidly and accurately while predicting new ones before they happen. LogicMonitor’s agentic AI solution, Edwin AI, has been generating an average $2 million annual savings per customer, leading to a meaningful uptick in recurring revenue. Vista believes that kind of performance stems from a guiding principle for AI deployment: Where AI-enabled solutions can deliver ROI for the end customer (not just the portfolio company), the odds of outpacing the market soar. 

Leadership from the center

At the heart of Apollo Global Management’s mobilization around AI is a center of excellence (CoE) set up to accelerate AI adoption across the portfolio. Unlike Vista, which is leaning heavily on internal resources, Apollo has staffed its CoE with two partners and an advisory board of external AI experts. This team has, in turn, built out a broad ecosystem of AI specialists, technology partners, and service providers. The firm has determined they are best in class at what they do, be it AI strategy, product development, sales and marketing effectiveness, or procurement.  

The CoE gives the diverse set of portfolio companies in Apollo funds a central resource dedicated to keeping them up to date on technology trends and working solutions. It appraises vendors, evaluates use case ROI, and generally creates an environment of continuous learning. At the same time, it serves as a go-to resource for portfolio company management teams across the funds. For companies with various levels of tech savvy, the CoE helps shape vision, set expectations, and introduce management teams to the appropriate implementation partner or partners, depending on what they need to learn or accomplish.

One of the CoE’s key functions is sharing what’s working across the portfolio and what has the highest potential for ROI. The CoE runs regular workshops with portfolio company management teams to demonstrate the art of the possible and lay out what’s getting results. The workshops begin with tangible AI success stories generating meaningful returns and end with homework assignments for portfolio company leaders. The firm asks each company to identify three to five potential use cases aimed at near-term strategic priorities and then develop a technology roadmap. Apollo has also created a structured playbook that includes an opportunity diagnostic as well as guidance on how to set up effective pilots and devise the strongest implementation plan.

Cengage, in which Apollo holds a minority stake, is currently executing eight AI projects to improve productivity in areas like sales enablement, customer care, content production, sales automation, and new product development. Early results are strong: Costs are down 40% in select content production processes, 15% to 20% via automated lead generation, 15% in customer care, and 10% to 15% in software development. Cengage has also launched two new AI products: Infosec Skills Navigator, which creates personalized cybersecurity training plans, and Student Assistant, a generative AI–enabled tutor bot that is in beta with hundreds of users.

Apollo’s Shutterfly, meanwhile, has been testing and implementing new AI capabilities across product development, software development, and customer care. The company launched a new AI auto-fill feature as part of its photo book creation path, which generated $5 million in new revenue in its first year. AI-enabled code assist has also produced productivity gains of 22% in an important replatforming project.

Specialization as leverage

Software specialist Hg is no less committed to using AI to generate value across the portfolio. But as David Toms, Hg’s managing director in charge of research, described on a recent episode of Bain’s Dry Powder podcast, the firm relies on the natural desire of portfolio company management teams to help each other coupled with their natural desire to compete, taking advantage of Hg’s tight focus on midsize business software companies with similar characteristics and operating models.

Sharing ideas is easier at Hg because its companies tend to face the same problems; a solution that works for one of them will often work for another. Consider how the firm is using generative AI across the portfolio to “refactor” code from outdated software languages to modern ones, extending the life of popular portfolio company products. Hg is also using generative AI to help management teams mine a massive database of companies around the world for sales prospects and M&A targets with specific characteristics.

By design, Hg looks for companies that operate in high-cost labor markets, where software spending is highest. It then breaks the business software market into clusters (accounting, payroll, industry-specific ERP solutions, etc.) that stand to add significant value by improving workflows to make those highly paid employees more efficient. This focus on efficiency serves as a bull’s-eye in thinking about how generative AI might help enhance value for companies and customers.

One of Hg’s accounting software companies, for instance, is using generative AI to take a significant leap forward in its customer proposition. An AI layer on top of its original software has completely transformed what the accountant sees upon opening the application every morning. Instead of providing raw numbers and a chart of accounts that need to be interpreted, the AI layer has already analyzed the data, figured out what needs to be addressed (based on historical patterns), and prioritized a clear set of actions. This gives the highly paid employee a jump-start on the day, freeing time for higher-value activities and boosting productivity.

Hg is pushing portfolio companies to innovate against their highest-value business priorities. Toms says the firm has taken a page from Silicon Valley’s book on innovation and encourages its portfolio company executives to explore potential generative AI use cases without a lot of preconceptions. “Don’t put an adult in the room,” he says, “because the adult will start saying, ‘Why don’t you do this with the blocks? Why don’t you try this?'" 

Instead, Hg identifies common challenges and encourages portfolio teams to solve them by playing on their natural instincts to help one other. When it comes to sharing ideas, Hg has found that management teams are more receptive to each other than to a firm-level managing director. As Toms puts it, “They want to talk to other people in the arena fighting the battle.”

Taking the next step

The moral of these stories is that there’s no one-size-fits-all approach to mobilization. The best model for tapping generative AI to create value depends on each firm’s culture, specialization, and resources. Every firm needs to solve for speed and focus, however. Powerful technologies like agentic AI (autonomous systems capable of goal setting and decision making) and “thinking” models are moving with blinding speed. These tools promise transformative impact for firms and portfolio companies that (a) understand them deeply and (b) have marshaled the resources to rapidly apply them to the strategic imperatives where they can change the game.

Are you ready? Here are several questions firms should be asking themselves:

• Have we assessed the risks and opportunities generative AI creates for each of our portfolio companies?
• Are our companies using generative AI to tackle their most important strategic priorities, or are they still dabbling?
• Have we thought through which approach to AI mobilization best suits our firm’s culture and resources? What combination of building a partner ecosystem, creating an internal team, directly supporting portfolio companies, and helping them collaborate most closely matches our playbook?
• Do we have the right expertise to help companies not only deliver the best technical solutions but also ensure that employees realize the technology’s benefits—that is, by understanding the change required, embracing it, and altering their ways of working?

Generative AI is no panacea. Like any technology, it is a tool that needs to be applied purposefully, practically, and strategically. And as with any technology, learning by doing is the key to harnessing its transformative potential. The time to get started is now.