Business Succession: The Five Phases of an Orderly Handover Process

Five phases are usually presented as a story of what comes next. That framing is useless the moment a process stalls, because a story has no failure condition. Treated as stage gates, the same five steps become decision points: each ends with a defined deliverable, and nobody proceeds until that deliverable exists and has been reviewed.

What a gate needs before it counts as passed

Four attributes belong in writing for every gate: the deliverable, the people who produce it, a realistic duration, and the observable signs that the work is not finished. The last carries the most weight. Processes rarely fail because a phase was executed badly, but because it was declared complete while its deliverable was still an intention.

Gate 1 — Readiness

Deliverable: a documented readiness position. Three cleaned financial years reconciled to the accounts, a normalised earnings figure in which every adjustment traces to a source document, a written statement of which functions depend personally on the owner, and a remediation list with named owners and dates.

Who works on it: the owner, whoever keeps the books, the tax adviser, and the people who will actually run the operational fixes.

Duration: the longest gate by a wide margin — one to several years where owner-dependence has to be genuinely reduced, and a few months of documentation work where the substance is already in place.

Not finished when: normalisations cannot be explained without the owner in the room; key contracts are undocumented or unsigned; customer concentration is known anecdotally rather than measured; the remediation list carries no dates. If the figures shift materially each time someone examines them, the gate is open.

Gate 2 — Value corridor and route

Deliverable: a written value corridor with the assumptions that produce its upper and lower bound, plus a decision on transaction route and target structure, with rejected alternatives and the reason for rejection recorded.

Who works on it: the owner and the transaction adviser, with tax input on structure. Valuation methodology itself is treated separately.

Duration: typically four to eight weeks once Gate 1 output is stable. It stretches whenever the underlying figures move, which is a Gate 1 defect surfacing late.

Not finished when: the corridor is a single number; the owner cannot state the price at which they would decline; the route decision rests on a conversation not yet held with a family member, a manager or a co-shareholder. An unspoken veto is an open gate, not a detail.

Gate 3 — Counterparty list and approach

Deliverable: a segmented, prioritised counterparty list with a documented reason for each entry; an anonymised profile and an information memorandum released for use; signed confidentiality agreements; a protocol governing who inside the company knows what.

Who works on it: the adviser runs outreach; the owner approves every name before contact.

Duration: typically two to four months from release to a workable set of indicative responses.

Not finished when: the list is short because research stopped rather than because the market is narrow; the memorandum asserts what the data room cannot support; contact began before names were approved. A leak here is normally a process failure, not misfortune.

Gate 4 — Letter of intent and due diligence

Deliverable: a signed letter of intent covering price mechanism, exclusivity period and conditions; a complete, indexed data room; a findings log in which every item is resolved, priced into the deal, or converted into a warranty or indemnity.

Who works on it: adviser, transaction lawyer, tax adviser, the counterparty team, and a small internal circle briefed under confidentiality.

Duration: typically three to six months, consuming most of the six to twelve months a transaction process usually runs.

Not finished when: findings sit open with no decision attached; the counterparty requests extensions without narrowing the issue list; financing is described as arranged but never evidenced. Repeated extension against a static issue list is the clearest measurable warning a process gives.

Gate 5 — Signing and closing

Deliverable: an executed purchase agreement, conditions precedent evidenced in writing rather than assumed, confirmed payment flows including any deferred element, and a closing memorandum listing what each side owes after completion.

Who works on it: lawyers on both sides, a notary where the structure requires one, banks, and the owner.

Duration: typically four to twelve weeks between signing and completion, driven by conditions precedent rather than by drafting speed.

Not finished when: consents from landlords, lenders or key counterparties are outstanding; obligations surviving completion have no named owner; funds flow is agreed verbally. What happens in the months after completion is a separate subject.

What breaks when a gate is skipped

Skipping never removes work — it relocates it to the most expensive point. Missing Gate 1 output reappears during due diligence as a price reduction, because a fact the buyer discovers is worth more to the buyer than the same fact disclosed voluntarily. Without Gate 2 there is no decline threshold, so every offer looks arguable. Without Gate 3 discipline, negotiation happens with whoever appeared first rather than the strongest fit. Without Gate 4 rigour, unresolved risk migrates into the warranty catalogue, where it stays live for years.

Which workstreams can run in parallel

Sequencing constraints bind decisions, not preparation. Legal tidy-up, data room assembly and document collection can begin during Gate 1 and continue throughout. Tax structuring can be modelled while the corridor is still being built.

What cannot be parallelised: no approach before corridor and route exist, no exclusivity before a value position exists, no signing before the findings log is closed. The rule is that anything producing input may run early; anything consuming a decision may not.

Abort criteria

Stopping and re-scoping is cheaper than continuing a process that has already failed. Grounds to stop: price revised downward twice on facts already disclosed; exclusivity lapsed once and extended without movement on the issue list; current trading diverged so far from the figures presented that the memorandum would need rewriting; a shareholder whose consent is required has withdrawn it; financing evidence promised and never produced.

Re-scoping means returning to the gate whose deliverable proved incomplete — usually Gate 1 — repairing it, and restarting from there. A process resumed with clean figures stands stronger than one dragged forward on the hope that the questions stop.

Frequently asked questions

What makes a phase complete?

The existence of its deliverable in reviewable form, not elapsed time. If the output cannot be shown to a third party and defended, the gate has not been passed.

Which gate fails most often?

Gate 1, though the failure usually becomes visible in Gate 4. Weak documentation is tolerable until a counterparty starts testing it under exclusivity, at which point it converts directly into price.

How long does the whole sequence take?

Twelve to twenty-four months from preparation to completion is realistic, with the transaction itself often six to twelve months. Preparation frequently consumes as much time as the transaction and determines its outcome.

Is a restarted process damaged goods?

Not if the reason for the pause is documented and remedied. A second run with evidenced figures and a defined decline threshold typically meets fewer price reductions than the first attempt.