The Integration

What this pack runs.

A synergy is a promise made in public before anyone owned it: cost removed or revenue gained a year, at run-rate, by a stated year, estimated before close by people who could not see inside the target and inherited by people who were not in the room. The premium is paid on day one and the number arrives over years, so the two never appear in the same document; by the second quarter after close the promise has been replaced in the boardroom by a percentage of initiatives actioned, and a percentage has no year in it. This pack starts the morning after close and turns the promise into arithmetic. What the deal promised a year, arriving in the promised year, counted for a horizon and discounted, is what the promise is worth. Against it: the premium once, the cost to achieve once, and the margin on the revenue the combination has cost. Divide the second by the first and you have the break-even synergy, the share of the promise that must reach the ledger for the deal to pay for itself, and its distance from the whole is how much of the promise can be missed. Then the date: the year the promise arrives at today's pace, read against the last year the promise pays, with the room between them in months. Nine sourced integrations hold the frame, three of them turning on a decision taken after close and six borrowed from deal records and labeled as such: two delivered the number and lost the base, three were written down or undone within five years, one carried a date and met it, one was raised to pay for a bid, one was taken through price, and one was made small and doubled after close. It stops on two halts: no one has reconciled what has reached the ledger to the accounts by line, and no one has counted what the combination has cost in revenue.

The casebooks these cases come from.

The pack reads nine sourced decisions through the framework. They are drawn from these casebooks, each one a set of verified records on a single decision type.

A verified record has passed a mechanical claim-versus-source check, a second model's read, and human approval. The count is what exists today, not a target.