The Integration Map: Five Phases, Five Ways to Wreck a Deal

Most people think the deal is done when it is signed. It is not. Signing is the easy part. The real work starts the day after.

Integration is five phases. Each one has its own way of killing your deal. I’ve watched good deals turn into a mess, and a few turn into something great. Here’s the map I wish someone had handed me the first time.

Phase 1: Due diligence

Everyone models the synergies. Almost nobody asks why the target’s clients actually stay, and whether the deal itself is about to remove the answer.

GE Capital found out the hard way. In 1995 it bought SOVAC, a French consumer credit house out of the Lazard world: discreet, relationship-led, wired into French business and politics. GE was the opposite. Direct, numbers-driven, American. It ran an operating model built for the US market, that did not fit a market built on local relationships.

SOVAC owned half of Credipar with Peugeot, and Peugeot stayed for one reason: Lazard was the guarantee. When GE took over, that guarantee was gone, and within three years Peugeot bought GE out. Every asset was still on the balance sheet. The thing that made them worth anything wasn’t, because it was never on the balance sheet to begin with.

The team that negotiates the deal is rarely the team that has to run it. GE’s dealmakers got the deal they modeled. GE’s operators inherited a relationship that had already ended, and didn’t know it yet.

In an integration, the hidden value is usually the first value you destroy. If diligence didn’t find it, you won’t know it’s gone until it already is.

And when it’s gone, the post-mortem will call it an execution problem. It wasn’t. Execution didn’t lose Peugeot. The decision to run Sovac the GE way lost Peugeot, before integration had even started.

Phase 2: Organization and governance

The question everyone actually has isn’t about strategy. It’s “who’s my boss, and do I still have a job.” Answer it slowly and your best people answer it for you, by leaving. They have options.

I saw this play out in a deal where one side took the lead: CEO, COO, and most of the key functional roles all went to people from the acquired business. Then came the headcount reduction plan. On paper, a social plan is designed to manage an exit cleanly and keep the business stable through it. In practice, it did neither. The people the company most wanted to keep were the ones with options, and they used them. They left for competitors, taking client relationships and knowledge with them.

None of that showed up on the integration reporting. It showed up a few months later as a business that quietly couldn’t do something it used to do well. The mechanism was simple: no one told the top performers where they stood, so the market told them instead.

Culture is the same trap wearing a softer name. Two companies means two sets of unwritten rules in one room. Steamroll one with the other without deciding to, and you haven’t saved anyone from a hard conversation, you’ve just broken the mechanism that made one of the two businesses work.

Phase 3: Communication

Silence isn’t neutral. It’s a decision, and other people fill it for you.

In another integration, system issues started hurting operations early, and then clients. Management knew and said nothing, not to the team, not to upper management. There was no explanation, no leadership presence. The CEO left the team to find out on their own. Morale collapsed first. People left second. The business started to collapse.

When I stepped in, I brought a whole team with me, a signal that this was being taken seriously. The first move was to tell the team what was broken, what we were doing about it, and what we needed from them. I followed up with regular employee meetings to keep them updated and answer their questions. That combination reversed the attrition and got the team back on board, before a single system issue was actually fixed.

You don’t need every answer to start talking. You need to say what’s known, what isn’t, and when you’ll be back. Then be back when you said you would.

Phase 4: Operations and systems

Someone recently told me firms like Accenture are excellent at integrating companies. They’re excellent at integrating systems. Not the same thing. And their business model has an opinion about what you need: program fees scale with the migration’s size and length, nobody bills by the hour for deciding not to migrate. The recommendation favors the recommender.

Most integrations reach for the ERP migration first, and that’s usually the mistake. On paper it looks clean: one org model, one platform, one way of working. The problem is applying a single model to different realities.

Understand why the businesses run differently first. It may be justified. Or not. Then simplify and harmonize the process before you touch the system, or all you’ve done is move the same processes and workarounds into a newer, pricier system. A common finance model and an integration layer between the existing ERPs often gets you most of the value without the migration’s cost or risk.

Whether a full migration is worth doing after that is an economics question, not a technology one. What does it create, what does it cost, can the business absorb the risk. That’s one of the hardest calls in the whole integration, and too many management teams let IT make it by default.

If the answer is yes, this is where the two businesses actually become one: shared data, one system, one way of working. It’s also where a bad cutover shows up immediately, as lost sales, delayed cash, and wrong stock, not as a line in a post-mortem. I’ve watched a cutover disrupt the exact operation the deal was meant to strengthen. ERP systems run the business. They are not an IT project, and the business needs to own the decision even when IT owns the delivery.

Phase 5: Synergies

Synergies are the whole point of the deal. The price paid assumes one plus one is worth more than two. That extra value comes in two forms, and most teams only chase one. Cost synergies are the easy half: shared functions, one head office, fewer vendors. They’re visible, they’re in the model, and someone usually does track them.

Growth synergies are the harder half: cross-selling into the other company’s client base, combined capacity that lets you take on business either firm would have turned down alone, a stronger offer to the market than either had on its own. That’s not efficiency for its own sake. It’s freed-up capacity put to work.

Synergies don’t show up because they were in the deal model. They show up because someone owns them after the deal closes, and almost nobody does.

Attention moves on the day the deal is announced complete. One-off integration costs keep arriving. The benefits that were the entire justification for the price paid arrive late, arrive partially, or don’t arrive. The deal was approved on a value creation case. That case doesn’t retire the day the deal closes. It has to be managed line by line until the value shows up in the financial statements, not before.

The map only works if you walk all of it

These phases don’t run one after another. They stack. You’re building the org chart while diligence is still closing. You’re talking to nervous employees before the new structure even exists. You’re merging systems before the synergy baseline is finished. That collision is exactly why the map matters: it stops you from pouring the whole team into whichever phase is loudest that week, while the quiet one in the corner is what actually kills the deal.

The deals that work aren’t the ones with the best model. They’re the ones where someone refused to let any of the five go silent.