A location consolidation is the decision to serve a site's customers from somewhere else. Closing a branch, merging two sites after an add-on, or folding a satellite into the site that was always doing the real work are the same commercial event: a book of customers whose service address or service route is about to change, and a share of that book that will not follow. The share that does follow is the number the case rests on, and it can be measured from data the platform already holds before the decision ships rather than discovered afterwards.
This guide is written for sponsors and operators running multi-site services platforms. It assumes the decision to close has already been taken on property, labor and contribution grounds, and it covers only the commercial half: how much of the book transfers, and what has to happen for it to. It is a different event from an acquired brand changing its name, which is covered in what the customer sees when you consolidate an add-on. There, everything the customer uses stays where it is and the identity changes. Here the identity can stay exactly the same and the place disappears, which no amount of notice repairs.
Most consolidation cases carry a retention percentage that nobody can source. Eighty percent of the book transfers, the model says, because eighty percent is the number that makes the case work. The share is not an assumption. It is largely determined by how far each customer now sits from the receiving site, and every platform already has the addresses.
Cut the closing site's trailing twelve months of revenue into three bands: customers inside the receiving site's existing service footprint, customers on its edge, and customers outside it altogether. Weight the bands by revenue and repeat rate rather than by customer count, because a closure does not remove an average customer. It removes a specific set of them, and the ones with the longest relationships are often the ones who chose that site for its location in the first place.
A platform that cannot produce this cut does not have a retention forecast. It has a hope with a decimal point. The cut takes a day of work and it changes the decision more often than operators expect, because the band outside the footprint is usually larger than the site's own managers believe.
Everything about a closure depends on who was doing the traveling.
The expensive mistake is applying the wrong model. A field-service closure treated as a communications problem produces a careful notice campaign aimed at customers who were never going to leave, while the revenue quietly goes to the jobs the receiving site declines to quote because the drive makes them unprofitable. A destination closure treated as a routing problem produces a plan to serve addresses that were never going to make the trip.
One table answers the question. For every customer of the closing site: address, trailing twelve months of revenue, job or visit count, date of last job, and drive time or route distance to the receiving site. Then three readings off it.
Write the resulting number down before the closure is announced, with the method attached to it. A retention share produced after the fact can be adjusted to match whatever happened, which is why the share agreed in advance is the only one that tests anything. The wider discipline of producing commercial numbers with their method attached is covered in the commercial evidence pack.
A platform can only keep what the receiving site can actually serve. This is the constraint that gets checked last and binds first.
Push a transferring book into a site already running near its practical limit and the retention plan becomes a service-failure event. Lead times stretch, the schedule slips, and the churn lands on the receiving site's own customers, who did not move, were not consulted and had no idea a consolidation was happening. That is the worst outcome available: paying for the closure twice, once in the book you lost and once in the book you already had.
So measure the receiving site's loaded capacity and current lead time before committing to a transfer share, and name the specific constraint. Chairs, bays, licensed staff, equipment or route hours all bind differently, and only one of them is usually the real limit. If capacity binds below the transferable share, there are three honest responses: move capacity along with the book, phase the closure so the transfer arrives in tranches, or accept a lower share and put the lower number in the plan. Quietly keeping the high number and hoping is not among them.
This is the mirror image of a site opening, where capacity is one of four inputs to a ramp that starts at zero and the risk is building too much of it too early. The opening case is covered in the new-location ramp curve. On a closure the capacity is already full of somebody else's customers, and the arithmetic runs the other way.
In a local services business the relationship frequently sits with a person rather than with the sign above the door. The stylist, the hygienist, the lead technician and the service manager who has been taking the same calls for nine years are the reason a meaningful part of that book is repeat business at all. In one five-brand consumer services platform we diagnosed, 38 percent of revenue arrived through untracked word of mouth, demand moving on personal recommendation that no system recorded. A closure puts that mechanism directly at risk, because word of mouth travels with people.
Which makes the sequence non-negotiable. Settle who is offered a role at the receiving site before any customer hears anything, and tell the staff before the customers. The human and legal reasons are obvious. The commercial reason is just as strong: a frontline that learns about the closure from a customer will not help you keep that customer, and the instinct to hold the announcement until the final week reliably backfires, because the people doing the work find out first and say so.
Then give the transferred book an owner at the receiving site. A book that arrives without one becomes background noise inside a larger site, and the general manager's number has to be restated to include it, otherwise the transfer is nobody's target and no miss is ever attributable. How company numbers get pushed down to the people carrying them is covered in setting sales quotas in a PE-backed company.
A closing site is rarely just a customer list. It has a backlog: booked jobs, deposits taken, memberships and maintenance plans with visits remaining, warranties on work already delivered, and service contracts with months left to run. These are commitments, and they are the highest-risk part of any closure, because a commitment missed in the same month as the closure is a complaint with an obvious cause and a reviewable story.
Handled well, the backlog is the best retention tool in the closure, because every remaining commitment is a scheduled reason to speak to a customer about where their service is coming from next.
A consolidation has a commercial line in the case, and it should be stated as a retained revenue share by distance band with a date by which the share becomes readable. Judging the result against the whole book is meaningless, because the band outside the footprint was written off on purpose. Judge it against the number that was named. For how the retention effect is measured on a cohort of affected customers, including why the clock starts at the switch rather than at the announcement, see the customer-facing side of a consolidation.
The result reports into the monthly commercial review alongside the rest of the plan, not into a project status report that closes when the doors do. The review's job on a miss is to tell a capacity problem from a distance problem from a handover problem, which is the whole reason the bands were separated in the first place; the exhibit set is covered in the monthly commercial operating review. The listing and local-search mechanics of a closed site sit with local search for multi-location platforms, and any rate change for transferred customers belongs to a pricing reset run on its own clock, well after the move.
The broader point is about symmetry. Platforms that add sites at speed also prune them, and one of our engagements grew its footprint 82 percent in 15 months, which means closures and mergers inside that footprint as surely as openings. Sponsors say operational work is where returns now come from: in S&P Global Market Intelligence's 2026 Private Equity Survey, 71 percent of general partners said they prioritize operational value creation over financial engineering, and 53 percent of limited partners said the same of the managers they back. A closure is operational value creation with the sign reversed, and it deserves the same evidence an opening gets. Where the fact base is thin, a commercial audit that maps customer geography, catchment overlap and site-level capacity establishes what the transferable book actually is before anyone commits to a date.
Have a revenue problem the board is asking about? Start a conversation.
There is no portable benchmark, and the published figures come from bank branches and retail stores rather than services businesses. The share is mostly a function of distance, so it should be computed rather than assumed. Cut the closing site's trailing twelve months of revenue into customers inside the receiving site's existing footprint, on its edge, and outside it, weight the bands by revenue and repeat rate, then read each band against the share of the receiving site's current customers who already travel that far. That produces a forecast from the platform's own book instead of a percentage chosen to make the case work.
Staff first, and before any customer communication goes out. Beyond the human and legal reasons, the commercial case is decisive: in a local services business much of the repeat book sits with specific people, and a frontline that hears about the closure from a customer will not help you keep that customer. Settle who is offered a role at the receiving site before the announcement, because the share of the book that transfers depends heavily on whether the people the customers deal with transfer too.
Against the retained revenue share that was named in writing before the closure, not against the whole book, because the band outside the receiving site's footprint was written off deliberately. Report it by distance band so a miss can be diagnosed: a shortfall inside the footprint is a handover or capacity problem, while a shortfall on the edge is the distance assumption being wrong. The result belongs in the monthly commercial review rather than in a project report that closes when the site does.
They are commitments and they set the closure date rather than following it. No closing date should land before the last booked commitment unless each job has been individually rescheduled with the customer's agreement. Deposits and prepaid plans transfer with the obligation named in writing from the receiving site, and that site also inherits warranty exposure on work it did not perform, so it needs the job records before a customer calls. A prepaid customer facing a longer journey is owed a choice, and a transfer with a credit usually costs less than a refund dispute.
No, and treating them the same is a common error. When an acquired brand is folded into a platform, the customer keeps the same site, the same crew and the same route, and what changes is the identity, so the work is mostly sequencing and communication. When a site closes, the identity can stay identical and the place itself disappears. Distance is a physical fact that notice cannot repair, which is why the first question is which customers are geographically keepable and the second is whether the receiving site has the capacity to serve them.