Summary of Key Takeaways

• Access to AI tools is becoming a baseline. Enterprise value comes from applying them to proprietary data, structured workflows, and specific customer outcomes.
• Measurable ROI is emerging in repeatable workflows and human augmentation, from software engineering to content repurposing.
• Change management is the biggest constraint. Capturing value requires clear governance, education, ownership, and redesigned workflows.
• Success should be measured by decision velocity: how quickly and effectively an organization can turn information into action.

Boardrooms and industry headlines are filled with enthusiasm for artificial intelligence, but for business leaders, the conversation has quickly shifted from potential to performance. Where is the commercial ROI? What creates a lasting advantage? And what can businesses confidently invest behind over a five-year horizon?

During a discussion at the AI & Tech Sandbox, leading private equity investors Laura Held, Partner, Shamrock Capital, and Lucy Dobrin, Managing Director, Providence Equity Partners, joined me to break this down. 

My core takeaway from the conversation is that as access to leading AI models becomes widespread, enterprise value increasingly depends on how businesses put that technology to work across their data, people, and operations.

‍

Cutting Through the Buzzwords: The Investor's Lens

"Initially for us, we were looking at how companies in our space are going to navigate AI... A lot of it was looking at what's the risk? Or what's the efficiency that these companies could drive with workflow, better speed, maybe better margins?" — Laura Held, Shamrock Capital

Investors start with practical questions: Can AI reduce friction, improve margins, speed up work, or create new revenue opportunities?

They are also looking more closely at what sits behind the growing number of AI claims.

Held shared, "There's a lot of buzzwords that companies will use to signal that they're on the right side of trends. Some have real substance. Some, when you dig in, don't really feel like an AI-native business or that they're using it in the right way."

That matters when investors are underwriting four- or five-year hold periods. Technology can change dramatically in that time, which makes enduring data advantages, customer relationships, workflows, and operating models increasingly important.

This reflects a foundational PMG thesis: Data is everywhere; intelligence is not.

Access to AI models and public data is increasingly common. The value comes from how an organization applies them to its own business.

‍

Fact vs. Fiction: Where Commercial ROI Sits

Some AI applications are already producing clear returns.

The Facts:

• Structured workflows: Software engineering, R&D, and data structuring are seeing strong returns because the work follows clear processes and rules.
• Content repurposing: Creative technology is helping brands adapt and reuse assets faster and at greater scale.
• Technology plus expertise: Companies such as VidMob, EDO, and Penta Group show the value of combining proprietary data and technology with human judgment.

The Fiction:

• "AI-powered" positioning without substance: Claims carry little weight without the data, workflows, or business model to support them.
• Standalone tools with weak differentiation: Products built on widely available models can quickly become interchangeable.
• Tool access as a strategy: Giving employees an open prompt box rarely changes a business on its own.

‍

The Real Constraint Is Coordination

"The learning and development piece and the change management is probably the number one thing we're spending time with our portfolio companies around right now." — Lucy Dobrin, Providence Equity Partners

The biggest challenge for many businesses sits inside the organization.

Buying software is relatively easy. Changing how teams work together, how decisions get made, who owns what, and where human judgment belongs takes much more effort.

As I said during the discussion:

"You can't just give people tools and say, 'Change.' You have to put a whole operational structure, governance, policies, and education in place."

Without that structure, teams face a blank screen: powerful technology with limited guidance on where to use it, which data to trust, or how its output connects to a business decision.

That is why AI increasingly becomes an organizational question. Businesses need to connect technology, data, strategy, activation, and measurement so information can move quickly into action.

‍

Building the Coordinated Enterprise

Three principles stand out for leaders:

• Lead by doing: As Lucy Dobrin advised, direct experience with the tools helps leaders understand where they create value.
• Augment human judgment: Technology and data become more valuable when paired with strong expertise and decision-making.
• Focus on decision velocity: Measure how quickly and accurately the organization can make consequential business decisions.

‍

So What Now? 

The next phase of AI adoption for businesses requires more than adding tools. We see five conditions for agentic enablement:

1. Vision & Outcomes: Decide what to build, what to automate, and where human judgment matters.
2. Operating Model: Establish the roles, ownership, and ways of working needed to move quickly.
3. Data & Governance: Create centralized, governed, transparent data that people and agents can trust.
4. Education + Enablement: Give teams the knowledge and support to turn access into practical capability.
5. Connected Ecosystem: Make clear choices about what to build, buy, and integrate.

All five should connect back to accountability for business outcomes and impact.

This is where we help businesses move from AI experimentation to enterprise enablement: bringing these five conditions together so technology, people, data, and decision-making work as one system.

That coordination is where the next wave of AI ROI will be won.