How Do You Integrate a Company Without Losing the Capability You Acquired?
How to integrate a company without losing the capability you acquired, using dependency mapping, staged operating-model choices and deliberate transfer.
By Stuart Andrews
The safest integration question is not where the acquired team reports. It is what capability the deal was meant to add, and what conditions allow that capability to keep working. Integration should define an end-state and an interim operating model, then change only what the evidence says must change while protecting the relationships and routines that created the value. This protects the capability you acquired while the combined business finds its shape.
Protect the capability you acquired during integration
I map capability before structure: customer trust, specialist judgement, delivery routines, data knowledge and informal networks. Each capability receives a protection decision, a transfer decision and an integration decision. Some should remain distinct for a period. Transfer needs a designed bridge: paired roles, shared reviews, documentation, shadowing and a point at which ownership changes.
Map capability before changing structure
Some should remain distinct for a period. The map exposes hidden dependencies. A product team may appear portable until the buyer sees the supplier relationships and decision context that make its performance possible. Retention bonuses alone do not transfer capability. They can keep people in place while the organisation removes the conditions that made them effective.
Where knowledge is held in practice
The map exposes hidden dependencies. The integration leader should publish what is changing, what is protected and what remains undecided. Silence turns uncertainty into attrition.
The handover risks leaders miss
A staged operating model gives leaders permission to preserve what works while they learn where the combined organisation should standardise.
A transfer rhythm for the first 100 days
Silence turns uncertainty into attrition. The acquired capability survives when the combined business can reproduce the judgement, relationships and routines that created the original value.
Signals that capability is moving
Silence turns uncertainty into attrition. The safest integration question is not “where does this team report?” It is “what capability did we buy, and what conditions let it keep working?”
Protect the people who carry context
McKinsey’s 2026 M&A guidance recommends defining an end-state and interim operating model quickly, while selectively transforming rather than changing everything at once.
The integration choice I would make
The safest integration question is not “where does this team report?” It is “what capability did we buy, and what conditions let it keep working?” I map capability before structure: customer trust, specialist judgement, delivery routines, data knowledge and informal networks.
Capability that survives the deal
Acquired capability survives when the combined business can reproduce the judgement, relationships and routines that created the original value. That is the standard I would use before changing the operating model.
Choose one decision that depends on the acquired capability and bring its new owners into the room. Put the evidence and constraint on the table, define what must be true for the choice and examine the effect on managers, customers, data, culture and the board.
Before changing reporting lines, I map the acquired capability as a chain of conditions. A specialist may need a particular customer relationship, a data convention and the freedom to make a judgement quickly. Remove one condition and the capability can appear to disappear even though the people remain employed.
The transfer map distinguishes knowledge from context. Documentation can capture a process, but it rarely captures why an exception was handled a certain way or which relationship makes a decision possible. Paired work and deliberate observation are needed for that context to travel.
An interim operating model is a protection mechanism, not a sign of indecision. It lets the buyer preserve a productive pattern while testing where standardisation will help. The integration leader should state the review date and the evidence that will inform the next structural choice.
Retention risk is often discussed as a list of names. I prefer a dependency view: which customers, decisions, systems or routines would be hard to reproduce if a person left? That view creates a transfer plan instead of a hope that loyalty will compensate for missing infrastructure.
The acquired team needs a credible answer to what will remain distinctive. If every difference is labelled temporary, people assume the buyer has already decided to erase the source of value. Specific protection commitments are more trustworthy than broad reassurance.
Integration succeeds when the combined business can reproduce the judgement and relationships that created the original performance, even after the original team is no longer operating as a separate unit.
The move I would make next: How Do You Integrate a Company
In the book I recount a local merger and acquisition at one of the world’s largest banks. The operational transition had already been attempted twice and failed. Multiple global partnerships, subsidiary teams and regulatory dependencies had to come together in the right sequence. The lesson was plain: retaining people is not enough when the operating conditions that let their judgement work have not been transferred.
Capability transfer needs a social route as well as a document. Pair a person from the acquired team with the person who will carry the work after integration, then let them solve a live problem together. The receiving colleague learns the judgement behind the method; the original expert sees which context has changed. I would review the pair’s decisions after thirty days and adjust the handover while the knowledge is still close. This is slower than copying a process note, but it leaves the combined business less dependent on one individual.
An acquisition integration needs a decision about what will not be standardised yet. Protect the local practice that keeps a customer promise or a specialist capability alive. Give it an owner and a review date. That is different from leaving the team alone. It makes the reason for the exception visible and gives the wider organisation a chance to learn before it copies the rule.
The first ninety days should include a few shared decisions rather than a long list of harmonisation tasks. Choose one customer issue, one operating control and one people decision. Let both teams show their reasoning, then record the rule that will apply next time. The record is more valuable than a slogan about integration because it can be tested.
People need a route to raise a concern when a new process loses important knowledge. Make the route part of the integration rhythm. Review the concern, decide whether the practice should travel and tell the team what happened. Silence is not alignment. It is missing information.
A combined organisation becomes stronger when it can say why a practice changed and what it kept. That explanation should be available to a new manager who was not in the deal room. It is a small test of whether the integration has become part of the operating system.
The measure I would watch is not uniformity. It is the quality and speed of decisions at the boundary between the two businesses. If people know who decides, where knowledge sits and when an exception will be reviewed, the integration is doing useful work.
The integration lead should keep a short list of decisions that remain open. Each line needs a person, a date and the evidence needed to close it. This prevents the deal team from treating every unresolved issue as a crisis, while stopping important questions from disappearing into a general workstream. Revisit the list with both businesses represented. When a local practice is retained, say what it protects. When it is changed, say what problem the change solves. That explanation is part of the capability being built. It lets managers make the same call later without waiting for someone who a
The people who join after completion should be able to understand the arrangement without the original deal documents. Put the decision rules, the protected capability and the review dates in a short operating note. Use a real example from the first months to explain where the rule helped and where it needed changing. That note keeps the integration from becoming a memory held by two or three senior people. It also gives a new manager permission to question a process without reopening the whole merger. Good integration is a living set of choices, and living choices need an owner who will look
The integration is complete enough to move when the work, not the presentation, tells the same story. Customers receive the promised service, specialists know where their knowledge sits and managers can make a decision without finding the person who negotiated the deal. Keep checking those conditions. They will change as the organisation grows.
Capability protection also needs an explicit owner during the transition. Ask which customer promise, technical practice, relationship or local judgement the acquired team carries, then identify the conditions that keep it visible while reporting lines change. Integration leaders should make those conditions part of the interim model, not leave them as informal goodwill. When the value is no longer needed, the organisation can choose to retire it knowingly rather than lose it through an accidental reorganisation.
Integration leaders should revisit the capability after the first operating cycle, not only at the formal close of the transaction. Ask whether customers still receive the same value, whether the team can make the decisions it used to make and whether the new structure has created unnecessary approvals. A small correction early is usually more respectful of the acquired capability than a late rescue after the strongest people have already left.
The strongest integrations preserve capability deliberately rather than sentimentally. If a routine no longer serves the combined business, change it and explain why. If a relationship or practice is still creating value, protect it while the wider model settles. The judgement is easier when leaders have named the capability, the evidence and the conditions that matter. Otherwise integration becomes a contest between structures, and the acquired value disappears inside the argument.
The acquired team should be able to say what must not be lost and why. That conversation is more precise than asking people to protect the culture in general. It might concern a customer insight, a technical judgement, a specialist relationship or a fast way of resolving exceptions. The integration team can then decide which conditions belong in the combined design and which were only useful in the old context. Naming the difference protects real capability while still allowing the organisation to change with intention.
That clarity gives the combined business a better chance of preserving what it bought while still becoming something new.
The integration plan should say how that learning will be heard. Create a route for acquired specialists to flag a loss of capability, give the integration owner authority to respond and review the evidence with the leaders who depend on the work. Without that route, concerns surface only after performance has fallen. With it, the combined organisation can adjust early and explain the reason for the adjustment.
That is how integration protects capability without freezing the organisation. It keeps the judgement that matters and changes the conditions that no longer serve the combined business.