Post-acquisition integration of two lenders

One mid-market lender acquired another. They were similar in size and different in almost everything else, and the deal closed with a synergy case built in diligence and a hundred-day plan built from convention rather than from the deal in front of it. Brand first, systems early, synergies everywhere. The acquired lender’s book was concentrated. Twenty broker relationships originated 60% of its volume, and three credit officers held most of what the book knew about itself.

The integration risk was not in the cost lines. It was in those twenty relationships and three people.

The pre-close review re-ordered the plan around where value could actually be destroyed. The conventional sequence would have spent the first hundred days on branding and core-banking migration while the brokers took calls from competitors. The First 100 Days Ledger inverted that order. Every broker and every named key person became a day-1–30 workstream with an owner, and the systems migration was scoped and then deliberately deferred to month 8, on the argument that a core-banking cutover cannot lose a broker relationship but an unreturned phone call can.

Protection ran first and it ran in person. The twenty brokers were visited inside thirty days, each of them with pricing and service commitments in writing, which is a logistics problem more than a strategy problem and was run as one. The three credit officers were retained by name before any structure was announced. A combined decision-rights ledger for the seam published on day one. Credit decisions kept dual sign-off for ninety days and were then unified under the merged credit policy.

The operating model merged in month 2–8. One credit policy, one origination process, one management cadence, each of them a drawn commitment with a frozen baseline. Cost synergies were released only once the protection metrics held, and they held, which is why revenue attrition came in at 1.9% against the 9.1% the data says convention delivers.

Synergies verified line by line against the deal model. One attribution line per benefit, and no double counting.