When to fire a legacy customer

The customer everyone finds difficult and the customer who is unprofitable are frequently different customers, and until you have computed it you are working from the loudest impression rather than from the ledger. Contribution margin at the account level — revenue less the variable cost of serving that account — is the right measure, and producing it usually requires the chart of accounts to have been rebuilt so that cost attaches to work rather than to a general bucket.

Then add the costs the ledger hides, because they are where legacy accounts go wrong. Rework and warranty call-backs. Unbilled visits that were absorbed to keep the peace. Emergency work quoted at standard rates. Staff time spent on scheduling exceptions this account requires. Payment behaviour, which is a real cost in a leveraged business: an account paying at ninety days when your terms are thirty is consuming working capital you are paying interest on.

Receivable behaviour deserves its own line in the analysis rather than a footnote. Pull the ageing by account, split beyond-terms balances into slow payment and genuine disputes, and note which accounts generate the disputes. An account that is marginally profitable on paper and habitually disputes invoices is consuming management attention that has a cost you can estimate even if you cannot book it.

What comes out is usually a distribution rather than a verdict: a handful of accounts carrying the business, a long middle at acceptable margin, and a short tail that is either underpriced or expensive to serve. The tail is the subject of this page, and almost every account in it is a pricing problem before it is an exit problem.

Legacy unprofitable accounts in an acquired business are rarely accidents. They are usually relationships: a customer the previous owner had known for twenty years, an arrangement made during a downturn and never revisited, a rate held because raising it would have meant a conversation the owner did not want to have. Practitioners who buy small businesses observe this pattern consistently — it is convention rather than a measured statistic, and we label it as one — and it means the tail is a residue of the seller's relationships rather than a failure of your team.

That has two consequences. The first is that the history matters and is only obtainable from the seller, which is another reason the pricing walk through the customer list belongs in the transition window while they are still available. Why this account is priced as it is turns out to be a specific story, and the story determines whether the account is worth repricing or worth exiting.

The second is that your team is watching. An inherited team that has served an account for years has its own relationship with it, and an exit handled badly reads as a statement about how you treat long relationships generally. The way through is not to avoid the decision but to make the reasoning visible: this is what the account costs, this is what we asked for, this is what they said.

Check the contract before doing anything. Confirm it was actually assigned at close, read the notice provisions, and look for the auto-renewal date, because in a surprising number of cases the exit is a matter of not renewing rather than of terminating. Concentration matters here too: an account inside your top-five revenue share is a different decision from one in the tail, regardless of its margin, because the cash gap is a different size.

The correct sequence is almost always to price the account properly and let the customer decide. Take them a number that makes the work worth doing, in writing, with a stated effective date and an explanation of what continuing gets them. A meaningful proportion accept — frequently more than the operator expects, because the account has been below market for years and the customer knows it. Those that decline have exited themselves, which is a materially better outcome than a termination letter, both for the relationship and for how the decision reads internally.

Where the gap is very large, phase it. An account fifteen years below current rates does not close in one step, and a stated multi-step path with dates is more likely to be accepted than a single jump that reads as punitive. Phasing also gives you a natural decision point: an account that accepts step one and refuses step two has told you something useful, and you have had two quarters of improved margin while learning it.

Attach an annual escalator to anything you re-paper. This is the whole reason the exercise is worth the effort — the account you fixed this year should not need fixing again in three years, and a contractual adjustment clause makes that automatic rather than conversational.

Two adjacent disciplines protect the result. Discretionary discounting has to be owned by a named person with a stated limit, or the increase granted at the front will be returned at the back within two quarters. And know the cases where the answer is to charge less rather than more — an entry tier designed to convert, a deliberate segment entry, a strategic account whose volume genuinely earns its rate. Those are decisions, not defaults, and they should be recorded as such so the next person to look at the account knows it was chosen.

Before you exit anyone, model the gap. Revenue leaves immediately and cost leaves slowly — you keep the technician, the truck and the lease for at least a quarter — so the short-term cash effect is worse than the margin arithmetic suggests. Put the departure into the thirteen-week forecast and check it against the debt-service calendar and your coverage headroom. A margin-improving exit that breaks a covenant is not an improvement.

The number worth having in front of you is how much revenue decline the business survives: contribution margin times revenue, less fixed costs, less debt service, solved for revenue. That is break-even re-derived for a leveraged operator, and it converts an abstract worry about losing customers into a specific figure. It is also the number that tells you how many exits you can run at once, which is usually fewer than the analysis makes you want to.

The exit itself is short, written, and gives real notice. State that you are unable to continue at a rate that works for both sides, give a date, offer to complete work in progress, and where you can, name an alternative provider who genuinely suits them better. Do not litigate the history. An exit conversation is not the place to explain that the account has been underpriced since 2011; that argument belonged to the pricing conversation and it was already had.

Then handle the internal version deliberately. Tell the team before the customer tells them, explain the reasoning in terms of the work rather than the personality, and make sure the account manager who has served them for nine years hears it from you first. And stage the exits — one at a time, watching cash — rather than clearing the tail in a quarter, because the analysis that identified the tail was produced from a chart of accounts you rebuilt recently and is worth a second look before you act on all of it at once.