From AI Theatre to EBITDA: A Better Model for Operational Due Diligence in M&A

AI is now present in nearly every deal thesis we see. It's framed as a lever for sharper targeting, deeper diligence, and smarter integration. But when we look at how value creation is actually landing inside private equity portfolios, AI is largely missing from the one metric that matters most: EBITDA.

Only 8% of PE firms describe themselves as "leading" with AI — meaning AI is materially impacting EBITDA and shaping the exit narrative, not just running in a pilot (Alvarez & Marsal, North America Value Creation Report 2026). The same report finds that 41% of respondents realized below 75% of planned value creation over the last 12 months. In European portfolios, 65% of firms achieved less than 50% of the value of plans developed over the past two years — a result that has stayed broadly consistent year over year.

At the same time, the deal math has gotten less forgiving. Bain's 2026 Private Equity Midyear Report puts it plainly: with purchase multiples and borrowing costs both elevated, a typical buyout now needs roughly 10–12% annual EBITDA growth to generate a 2.5x return over a five-year hold. In the 2010s, 5% growth was often enough. Bain's shorthand — "12 is the new 5" — isn't a slogan. It's a constraint every sponsor is underwriting against right now (Bain & Company, Private Equity Midyear Report 2026).

We see this tension up close, deal after deal. AI is in the narrative. EBITDA impact is not.

The real problem isn't the tools — it's the workflow upstream of close

Across our work with sponsors, operating partners, and portfolio executives, one pattern repeats. Synergy cases are built on fragmented ERP and spend data, inconsistent item masters, and management-supplied spreadsheets. Value creation planning rarely happens before the letter of intent, even though most respondents believe earlier planning would improve outcomes. AI gets deployed in the safest domains — FP&A, dashboards, document review — rather than the procurement, supply chain, and integration workflows where synergies are actually captured. And management bandwidth, cross-functional coordination, and change resistance consume the time that should go toward execution, because baselines still need to be reconciled after close.

The result is familiar to anyone who has sat in an integration steering committee. Revenue volume growth is the single largest plan shortfall, cited by half of respondents — twice as often as any cost-side category. Technology enablement, change management, and talent gaps follow close behind, which tells us the binding constraint is execution capacity, not thesis quality.

The academic record backs this up at scale. Lev and Gu's The M&A Failure Trap documents a 70–75% failure rate across roughly 40,000 acquisitions — a rate that has been rising, not falling (CFA Institute, Book Review: The M&A Failure Trap).

From where we sit, the issue isn't whether firms have access to AI tools. It's whether operational due diligence has been redesigned so those tools are pointed at the right thing — normalized data and governed workflows, upstream of close — instead of layered onto the same fragmented foundation that has always produced disappointing synergy cases.

Our view: diligence should produce an executable baseline, not just a report

At Vantage Pointe Partners, we've built a five-phase, operator-led, AI-enabled model for operational due diligence. The core conviction is simple: diligence should produce a normalized operational baseline and an executable synergy case that carries directly into post-close workflows — not a static report of estimates and observations that gets shelved the day the deal closes.

Phase 1 — Data mobilization and visibility

We extract and normalize ERP, item master, spend, and supplier data across buyer and target before any synergy estimate is developed. Talent capability and data quality are consistently the top two barriers firms cite when trying to scale AI across a portfolio — resolving them here changes every output downstream.

Phase 2 — Current-state assessment

We use that normalized data to build a decision-ready view of actual spend by category and supplier, network footprint across plants and distribution centers, and transportation cost-to-serve by lane. This becomes the operating baseline the deal is actually underwriting, not the one management described in a board deck.

Phase 3 — Scenario modeling and synergy quantification

With clean baselines in place, we deploy AI-enabled tools to model supplier overlap, network rationalization, and resilience trade-offs. We treat AI as infrastructure for scenario modeling, not a dashboard layer — because the underlying data is normalized, the outputs are scenarios with traceable assumptions, not educated guesses dressed up as analytics.

Phase 4 — Executive alignment and synergy case design

We convert scenario findings into a defendable synergy case with explicit category owners, milestones, and KPIs traceable back to the Phase 2 baseline — a base, upside, and downside case ready for investment committee underwriting.

Phase 5 — Implementation support and KPI tracking

We carry the diligence baseline into the real operating environment: standing up governance routines tied to pre-close KPIs, supporting supplier negotiations, and protecting leadership bandwidth against the coordination friction that derails so many integration plans.

Our documented client results across procurement transformation and network optimization engagements reflect OpEx savings in the 5–25% range and working capital improvement in the 7–15% range, depending on sector and starting baseline. These are client-documented outcomes from our own engagements, not third-party benchmarks.

What this means for sponsors, CFOs, and portfolio executives

For sponsors and operating partners, synergy cases move from estimate-based to evidence-based, and exit narratives shift from "we think" to "we delivered" — auditable against pre-close baselines rather than reconstructed after the fact.

For CFOs and CPOs, the most common miss — revenue volume growth — gets tackled with real visibility into cost-to-serve and supply chain constraints, and pricing decisions get built on normalized data instead of proxies.

For portfolio executives, the first 100 days get spent executing against an already-agreed synergy case instead of rebuilding the baseline from scratch under time pressure — which is exactly when management bandwidth is most scarce and most valuable.

Pressure-test the next deal before close

If the next deal's synergy case depends on assumptions that can't be traced to normalized ERP data and executable workflows, the question isn't whether your firm has adopted AI. The question is whether diligence has been redesigned to capture value in the real operating environment — not the one described by management-supplied inputs.

Most PE respondents believe pre-LOI value creation planning would improve outcomes. Far fewer actually practice it. That gap — between what sponsors know would work and what most diligence processes actually deliver — is the defining opportunity in operational M&A today.

We built this model to close it. If you're preparing for an upcoming transaction and want to pressure-test the synergy assumptions before close, we'd welcome the conversation.

Explore our M&A Advisory & Synergy Value Creation practice

The five-phase operational due diligence model described above is the core of our M&A practice. See how it connects diligence to execution — from data mobilization through post-close KPI tracking — on our M&A Advisory & Synergy Value Creation service page. For a complementary read on what operational diligence catches that financial diligence misses, see What Operational Due Diligence Actually Catches.

Sources

• Alvarez & Marsal, North America Value Creation Report 2026
• Bain & Company, Private Equity Midyear Report 2026
• McKinsey & Company, 2026 M&A Trends: Navigating a Rapidly Rebounding Market
• Lev, B. & Gu, F., The M&A Failure Trap (Wiley, 2024) — reviewed in CFA Institute, Enterprising Investor