By Bryan Knipe, founder of TriPinnacle Advisors
Acquisitions that disappoint usually fail in one of two ways. Either the deal was decided on a story, a strategic fit everyone liked, without a business case anyone could hold it to, or the deal was sound and the integration was treated as something to work out after closing. Both are avoidable, and both are cheaper to avoid before signing than after.
At Optum I evaluated acquisition targets, including AI and analytics companies, built the integration business plans and took part in diligence and negotiations. I also led the evaluation of buying a provider's cost-accounting intellectual property, a buy-versus-build decision, and later led the transition when a large health system moved its entire analytics organization to us, which is an integration in everything but name. These are the steps I follow.
1. Ask whether to buy at all
Before evaluating a target, test the alternatives: build it yourself, partner, license, or buy something smaller. Buying makes sense when it gets you something you can't build in time or at a reasonable cost: customers, a team, data, technology or a regulatory position. If the answer to "why buy?" is mainly speed, put a number on how much the speed is worth.
2. Write the investment thesis in one paragraph
What exactly are you buying, what will you do with it that the seller couldn't, and what will it be worth to you in three years? Every later step tests this paragraph. If diligence breaks it, walk away, even late.
3. Commercial diligence
Talk to the target's customers, not just its management. Are they happy, are they staying, and would they buy more if you owned it? Look at revenue concentration, contract terms, renewal history and the pipeline. The customer calls usually tell you more than the data room.
4. Operating diligence
How does the business actually run? Who does the work, on which systems, at what cost, and which few people would take the business with them if they left? Map the processes that matter to the thesis. This is where the integration cost hides.
5. Technical and AI diligence
For technology and AI companies, test the claims directly:
- Does the AI work on your data, not just the demo set? Ask for a test on a sample you provide.
- Who owns the data and the rights to use it? Models trained on customer data the target can't transfer may be worth much less to you.
- What does it cost to run? Compute and model costs at your volume, not theirs.
- Who maintains it? If two engineers understand the system, the deal needs a plan to keep them.
- Security and compliance: especially with health data, how is it protected, and what would your Legal, Compliance and Information Security teams need to see?
Having built a production AI platform myself, I look past the demo to the parts that decide whether it survives in production: data handling, output checks, cost control and tests.
6. Build a business case you can hold the deal to
The business case should show what the combined business earns with and without the deal, the synergies by type and timing, the integration cost and the one-time costs, in scenarios. Cost synergies are easier to deliver than revenue synergies; discount the revenue ones accordingly. Name an owner for every synergy, so a year later someone can say whether it happened.
7. Plan the integration before you sign
The integration plan should exist before the deal closes, with decisions made on the big questions:
- What stays separate and what gets combined: brand, systems, teams, processes.
- Who leads the combined business and who are the people you must keep, with a plan to keep them.
- What happens on day one for customers and employees, and in the first 100 days.
- How you keep the business running during the transition. In the Optum transition, keeping the health system's reporting working throughout was the measure that mattered most.
8. Measure it afterward
Track the deal against its business case for at least two years: revenue, synergies, integration cost and the people who stayed. It is the only way an organization gets better at acquisitions.
Where I help
As a fractional strategic CFO or interim COO, I evaluate targets, run the operating and AI diligence, build the business case and integration plan, and lead the integration after closing.