The PE AI playbook: how portfolio operators deploy AI across multiple franchise brands

The 7% problem

The numbers tell a strange story. According to FTI Consulting's 2026 Private Equity AI Radar, 95% of PE funds report that their AI initiatives are meeting or exceeding the original business case. By most measures, that's a success rate any operating partner would celebrate.

The other 93% are running pilots, point solutions, or single-function tools that work at one brand but don't connect to anything else. FTI found that 36% of portfolio companies use AI across multiple use cases, but almost none have built the connective tissue to make those use cases work together or replicate across brands.

For PE firms managing franchise portfolios of 5 to 30+ brands, this is the central question: how do you move from 30 separate vendor evaluations to one playbook that compounds across the portfolio?

The three modes of portfolio AI

BCG's framework for enterprise AI adoption maps cleanly to how franchise portfolios operate. Most portfolio companies sit in one of three modes:

Deploy is where roughly 60% of companies start. Hand out tool licenses. Let each brand figure out its own AI stack. A plumbing franchise picks one scheduling tool, the HVAC brand picks another, the cleaning brand experiments with a third. Each reports modest productivity gains of 10-15%, per BCG's estimates.

Reshape is where the operating value sits. Instead of optimizing individual tasks, the portfolio operator restructures how specific functions work across brands. FTI Consulting reports EBITDA improvements of 5% to 25% from AI deployed across operations and back-office functions. The difference from "deploy" mode: the AI changes the workflow, not just the speed of the existing one.

Invent is rare: new revenue streams or business models enabled by AI. Few franchise portfolios are here yet.

The problem is structural. FTI found that 40% of PE firms manage AI investments at the portfolio company level using decentralized models. Each brand runs its own evaluation, its own vendor relationships, its own integration work. And 36% of firms that do have an AI strategy lack specific milestones or KPIs for measuring whether it's working.

What a replicable playbook looks like

Apollo's approach to portfolio AI, documented in MIT Sloan Management Review, offers the clearest public example. Rather than broad digital transformation programs, Apollo's portfolio operations team identified 2-3 high-impact use cases per company and built dedicated capability around each.

At Cengage, one of Apollo's portfolio companies, this meant targeting four specific functions: content production (40% cost reduction), lead generation (15-20% improvement), customer care (15% improvement), and software development (10-15% improvement). Eight AI projects, each tied to a measurable operational outcome.

The pattern that matters for franchise portfolios is the sequence, not the specific numbers: pick a function that exists across multiple brands, prove the approach at one brand, measure the result, then replicate.

For a home services portfolio operating plumbing, HVAC, and electrical brands, that function might be after-hours call handling. The workflow is nearly identical across brands: a customer calls after 5pm, needs service, and gets routed to voicemail or a third-party answering service. Prove that AI call routing works at the plumbing brand. Measure the booking rate change. Then deploy the same approach at HVAC and electrical with brand-specific configuration rather than brand-specific evaluation.

Why decentralized AI fails franchise portfolios

The franchise model adds a layer that general PE portfolio management doesn't face. Each brand already has a franchisor-franchisee dynamic where technology adoption requires buy-in at the location level. When the portfolio operator adds a third layer of decision-making on top, the adoption challenge compounds.

A decentralized approach means Brand A's franchisees learn one system while Brand B's franchisees learn a different one. The portfolio operator can't compare performance across brands because the tools measure different things. The data from Brand A's customer interactions lives in one system while Brand B's lives in another.

FTI Consulting's research suggests the alternative: PE firms that centralize AI orchestration and build shared capabilities across portfolio companies capture more value than those running distributed experiments. The firms doing this well standardize the evaluation criteria, the measurement framework, and the deployment methodology rather than standardizing every tool, then let each brand configure within those guardrails.

The franchise-specific advantage

Franchise portfolios have something most PE portfolio companies don't: operational repetition across hundreds or thousands of locations. A 200-location HVAC franchise runs roughly the same workflow at every location. A portfolio of six home service brands runs variations of the same five or six core workflows: inbound call handling, scheduling, dispatch, invoicing, follow-up, and review generation.

That spending is happening whether the portfolio operator coordinates it or not. The question is whether each brand's technology investment produces isolated gains or contributes to a portfolio-wide capability that gets stronger with each deployment.

FRANdata's 2026 outlook predicts a shift toward agentic AI systems capable of real-time decision-making and system-wide optimization across franchise networks. For a single brand, that's ambitious. For a portfolio operator who has already proven the approach at three brands and has the data to show what works, it's the next logical step.

Building the playbook

The franchise portfolios that will pull ahead in the next 18 months are the ones that stop treating AI as a brand-level technology decision and start treating it as a portfolio-level operational capability.

That means an operating partner who owns the AI playbook across brands. It means a measurement framework that lets you compare call handling performance at the plumbing brand against the HVAC brand against the electrical brand. It means vendor relationships negotiated at the portfolio level, not the brand level. And it means deployment methodology that accounts for the franchisor-franchisee dynamic at every brand.

The 7% who have reached enterprise scale didn't get there by running 30 pilots. They built something that works, proved it, and replicated it.