AI Operating Model for PE Portfolio Companies: Operations First, AI Second

Every PE sponsor I talk to wants AI in their portfolio companies. The board decks mention it. The value creation plans include it. The CEOs are under pressure to show an AI strategy. And most of them are approaching it exactly backward.

The standard playbook goes like this: hire an AI consultancy, identify use cases, run a pilot, measure adoption. It sounds logical. It almost never works. Across the PE-backed companies I've led and advised, the failure rate on AI initiatives that follow this playbook is staggering. Not because the technology doesn't work — but because the organization isn't ready for it.

AI amplifies whatever operating system it's deployed into. If your decision rights are clear, your workflows are documented, your data flows are clean, and your teams are aligned — AI will compound that efficiency dramatically. But if your operations are broken — unclear ownership, siloed data, inconsistent processes, misaligned leadership — AI will amplify that dysfunction just as efficiently.

This is the fundamental mistake most PE portfolio companies make with AI. They treat it as a technology initiative when it's actually an operations initiative. The technology is the easy part. The hard part is getting the operating model into a state where AI can deliver value.

Here's what an AI-ready operating model looks like. First, decision rights are clear. Every process that AI will touch has an owner who can define requirements, approve changes, and be accountable for outcomes. Without this, AI projects stall in committee. Second, data flows are unified. AI needs clean, accessible, connected data. If your portfolio company runs on spreadsheets, tribal knowledge, and disconnected systems, no amount of AI will fix that. Third, processes are documented and consistent. AI automates what exists. If what exists is different in every region, every team, or every customer segment, AI can't standardize it. Fourth, the leadership team is aligned on where AI adds value. Not 'we need an AI strategy.' Specific use cases with measurable ROI tied to the value creation plan.

The approach that works in PE portfolio companies is two-phase. Phase one: fix the operating model. Install decision rights, clarify ownership, build cadence, unify data flows, standardize processes. This takes 60 to 90 days with the right operator in the room. Phase two: deploy AI into the clean operating model, starting with the highest-ROI use cases — typically in customer operations, professional services delivery, sales intelligence, or financial reporting.

We've deployed this approach at a PE-backed managed services company with $150 million in revenue and flat margins for three years. The previous AI initiative had stalled at 12 percent adoption. We spent the first 60 days fixing the operating model — redesigning delivery workflows, clarifying ownership, installing weekly cadence. Then we deployed AI-first workflows into the clean processes. Adoption went from 12 to 67 percent. Annual savings: $2.4 million.

The difference wasn't better AI. It was a better operating model underneath the AI. The same technology that failed at 12 percent adoption succeeded at 67 percent — because the organization was ready for it.

For PE sponsors evaluating AI strategies in portfolio companies, the diagnostic question is simple: is the operating model ready? If the leadership team can't tell you who owns each major process, if data sits in disconnected systems, if decision-making requires CEO escalation — AI will not fix those problems. It will automate them.

Operations first. AI second. That's not a conservative approach. It's the only approach that delivers sustainable ROI in a PE timeline. Most advisory firms can do one or the other. We do both — because you can't separate AI strategy from operating model strategy. They're the same thing.