Ranking customers by revenue tells you very little about which ones are worth having. Most owners know this and rank them by revenue anyway, because it is the only number that is easy to get.
Here is a comparison that came up in a session recently. The owner had one account paying around $3,000 a month that required about two and a half phone calls a week, worked only within a narrow Sunday window, generated regular callbacks and was hard on his staff. He had other accounts paying roughly double that which he had not spoken to in six months.
He wanted to fire the first one. His stated reason was that they were rude and difficult. That reason was true and it was not the useful part.
The useful part is that he had never priced the two and a half calls a week. Once you do, you get a different conversation than “do I like this customer,” and usually a different decision too. He ended up keeping that account.
This article is the audit I run. Four dimensions, a scoring method and, more importantly, the order to do things in once you know the answer. The order is where most of these go wrong.
Revenue is the wrong sort
An account’s revenue tells you what it pays. It tells you nothing about what it costs, and the largest costs of a difficult account never appear on an invoice.
Four things determine whether a customer is worth having:
Profit. What is left after the direct labor, supervision, travel, administration, payment processing, rework and callbacks that account actually consumes. Not gross revenue and not your blended margin.
Strategic value. Does this account produce referrals, geographic density, credibility with a segment you want or capability you can resell? Some unprofitable accounts are genuinely worth keeping. Most are not, and owners are inconsistent about which is which.
Risk. Concentration, liability exposure, payment reliability and how much of your business disappears if this one relationship ends.
Management burden. The hassle factor. How much of your attention and your team’s attention the account consumes, and whose attention specifically.
Revenue correlates with none of these reliably. That is the entire problem.
Dimension 1: profit
Start here because it is the only one with a real number, and because owners are usually wrong about it in a specific direction.
For each of your top accounts, calculate:
Revenue
− Direct production labor
− Supervision and coordination time
− Travel and drive time
− Rework, callbacks and warranty work
− Payment processing and collections effort
= ContributionTwo things to be strict about.
Price your own time into it. If you are personally handling this account’s problems, that is a cost. Most owners leave it out, which makes difficult accounts look far more profitable than they are. If you have not calculated your own hourly rate yet, do that first, because it is the input that makes this honest.
Count rework properly. An account generating callbacks is consuming labor twice and often at premium urgency. A job redone once has roughly half the margin you think it does.
You do not need perfect cost accounting. Directionally accurate is enough to change a decision, and directionally accurate is achievable in an afternoon.
Dimension 2: strategic value
This is where owners talk themselves into keeping things.
The honest question is not “is this account strategically valuable.” It is: am I keeping this because it is strategic, or because I am afraid of replacing the revenue?
Those feel identical from the inside. A test that separates them: if this account paid the same amount but produced no referrals, no density and no credibility, would you still keep it? If yes, you are keeping it for the revenue and the strategic case is decoration.
Real strategic value is specific and you can say exactly how it works:
- It produces a measurable number of referrals per year
- It anchors a geographic cluster that makes surrounding work more profitable
- It is a reference account that closes similar buyers
- It teaches you a capability you intend to sell more of
“They have been with us since the beginning” is not strategic value. It is history, and history is a reason to be gracious about how you make a change rather than a reason not to make it.
Dimension 3: risk
Two kinds, and owners systematically underweight both.
Concentration
The number that matters is what percentage of revenue a single account represents, compared against your operating margin.
An owner I worked with had a single account at roughly 12% of revenue while running an operating margin in the high teens. They described it as their best account, which was true. It was also, in practical terms, most of their profit. If it left, the business would not lose 12% of its comfort. It would lose most of its cushion.
That is not an argument for firing it. It is an argument for knowing it, contracting it properly and building a replacement pipeline before the question becomes urgent rather than after.
My general rule: any single account above 10% of revenue gets a written agreement, a named relationship owner other than you and a documented plan for what happens if it ends. Above 20% and I would treat replacement as an active project regardless of how good the relationship is.
Liability
This one is less visible and occasionally larger than the account itself.
A different owner was doing roughly six figures a year of work for a multi-site institutional client on a handshake. No contract, no defined scope, no liability language. Their own words about the exposure: if someone slipped on a floor their team had just cleaned, the claim goes to the property owner and the property owner comes to them, and there is nothing on paper defining what he agreed to.
If you are doing meaningful volume with no signed agreement, that is a risk finding independent of anything else in the audit, and it does not resolve by itself.
Dimension 4: management burden
The hassle factor. Hardest to quantify and usually the reason the audit was started.
Score each account on five things over a typical month:
| Contacts per month | Calls, emails and texts requiring a response |
| Whose attention | Owner, manager or coordinator. Weight owner attention heavily |
| Schedule rigidity | Narrow windows, weekend requirements, short-notice changes |
| Staff impact | Do your people dislike this work, and does it affect retention |
| Payment friction | Chasing, physical collection, disputes, slow terms |
Then convert to a number. Total hours a month, priced at the rate of whoever spends them. Two and a half owner calls a week at twenty minutes each is roughly three and a half hours a month. At an owner rate around $70 an hour that is $245, plus the fragmentation cost of the interruption, which is real and larger than the time itself.
That is the calculation almost nobody runs, and it is the one that turns “this customer annoys me” into a number you can put next to the contribution figure.
What the burden number actually tells you
Here is the part that surprises people. A high hassle number is frequently not evidence that the account is bad. It is evidence that the account is unmanaged.
Look at the components. Narrow scheduling windows are a scheduling policy problem. Payment chasing is a terms and collections-ownership problem. Callbacks are a documentation and standards problem. Contacts routing to the owner is a delegation problem.
Almost none of those are properties of the customer. They are properties of how you are running the account.
I have watched owners fire a difficult account, replace it with a similar one and produce the identical situation within a year, because the thing generating the hassle travelled with them. Your capacity to organize, prioritize, plan and execute is what makes an account run well. That is mostly not the customer’s contribution.
Put differently: if you can run your hardest account cleanly, you have built a template you can run on a hundred more. If you cannot, firing this one buys you a quiet quarter.
The decision grid
Score each account on contribution and burden, then place it.
| Low burden | High burden | |
|---|---|---|
| High profit | Protect. Contract it, name an owner, do not touch the price. | Fix. This is where most of the value is. See the sequence below. |
| Low profit | Reprice. Usually the easiest wins in the whole business. | Reprice hard, or release. |
Two notes on reading this.
The high profit / high burden box is where the money is, and it is the box owners want to fire from. The account is already paying well. The hassle is usually about how the account is run, and that is fixable. Fixing it is almost always worth more than replacing the revenue, because replacement costs acquisition spend and a ramp period while the fix costs a few conversations.
The low profit / low burden box is the one that gets ignored. Nobody thinks about quiet, mediocre accounts, which is exactly why their prices have not moved in three years. An owner I worked with sent twelve emails raising prices between $25 and $100 on accounts like these. Every one accepted. That was roughly $800 a month found in an afternoon, with no acquisition cost and no delivery change.
If you do one thing from this article, do that one. It is the highest return per hour of anything in this piece.
The sequence
This is the part that matters more than the analysis, and it is where I see good decisions turn into damaged relationships.
If an account needs a price increase, a written agreement and better payment terms, do not do them in the same conversation. I have watched an owner attempt all three in one week and convert a routine adjustment into a full renegotiation of the relationship, including terms that had not been in question.
The order I would generally use:
1. Price first, in pieces
Start with the specific line items you know are underpriced rather than an across-the-board increase. Smaller, more defensible, easier to say yes to.
Tradeoff: this is slower and you will leave some money on the table in the first round. You are buying a much lower chance of triggering a competitive bid.
2. Agreement second, framed as an operational requirement
Once pricing is settled, introduce the written agreement, and introduce it as something that is happening rather than something being negotiated. Insurance and liability review is the honest and usable frame, because it is generally true and because it moves the agreement from “Andrew wants something from you” to “this is a requirement we both have to satisfy.”
Bring the existing terms, written down, unchanged. The agreement should document the deal you already have. If you change price and paper the deal at the same moment, you have reopened everything.
Tradeoff: you are spending a conversation on something that produces no immediate revenue.
3. Payment terms third
ACH, autopay, whatever removes the friction. Last, because it is the smallest item and the easiest to concede if you need something to concede.
4. Relationship owner last, and permanently
Move day-to-day contact off yourself. This is what stops the hassle from regenerating.
Tradeoff: the account will be handled differently than you would handle it, and the first few months will include mistakes you would not have made. That is the price of the hours.
A note on institutional customers
Owners routinely over-read the emotional temperature of large institutional clients. You are dealing with people executing a role. The person who cuts checks will cut them however you ask. The person reviewing agreements cares whether the agreement exposes their employer, not whether you asked for one.
The friction you are anticipating is usually not there. Be straightforward, have standards and make yourself easy to work with procedurally, and most of these requests are processed rather than debated.
The thing underneath all of this
There is a reason the audit so often produces “this account is fine, I just cannot stand dealing with it.”
A lot of what reads as customer difficulty is the owner’s own operating anxiety. Waiting on a check is not expensive if you have reserves. It is only expensive if the check arriving late means a hard conversation about payroll.
I built an entire collections process once around my discomfort with waiting to be paid. It would never have survived twice the volume, it was worse for customers and it took me about a year to understand that I had solved the wrong problem. The problem was the size of the cash reserve, not the speed of collections. Once the reserve was right, the same late payments stopped mattering.
So before you fire anyone, ask two questions:
How much of this goes away if the business holds two months of operating expenses in cash? Usually most of the payment-related burden and some of the scheduling rigidity, because you stop needing every job to land on time.
How much goes away if someone other than me owns this relationship? Usually most of the rest.
If the answer to both is “most of it,” the customer was never the problem, and replacing them will not fix anything.
Then who should you actually fire?
Some accounts should go. In my experience the ones that genuinely should are:
- Negative contribution after honest costing, where the price cannot be moved enough to fix it
- Accounts that damage your ability to keep staff. People are harder to replace than revenue, and an account that burns good employees is expensive in a way that never shows up on the account
- Accounts with unresolvable risk, where they will not sign anything and the exposure is real
- Accounts that conflict with what you are building, where you would not want more customers like this one
That last category is legitimate and I would not talk anyone out of it. If an owner tells me they will not do business with a certain kind of client, that is a decision about what the company is, not a spreadsheet output.
One condition though: know what it costs before you do it. “This account is 8% of revenue, it contributes roughly this much margin, and I am choosing to give that up because I do not want to build a company around clients like this” is a sound decision. The same choice without the numbers is a mood.
Do the audit
- List your top fifteen to twenty accounts by revenue
- Calculate contribution for each, with your own time priced in
- Score management burden across the five components, in hours, by whose hours
- Place each on the grid
- Start with the low-burden repricing. It is the fastest money in the business
- Then take your highest-value high-burden account and run the sequence on it as a template
Download the Customer Audit Scoring Sheet →
Run it annually, or after any quarter where a single account has consumed a disproportionate amount of your attention.
Frequently asked
How do I calculate contribution if my bookkeeping is not job-costed? Estimate. Take the hours the account consumes in a month, price them at your loaded labor rate, add materials and a share of coordination time. Directionally accurate is enough to sort accounts, and sorting is the goal.
What if my biggest account is also my most difficult? Common, and it is the fix box rather than the release box. Work the sequence. Do not start by threatening the relationship, and do not start the conversation until you know what losing it would cost.
How many customers can I afford to lose in a price increase? Depends on contribution, not revenue. If an account contributes 30% margin and you raise prices 10%, you can lose roughly a quarter of that revenue and end up level on profit while freeing capacity. Run your own numbers before you decide how brave to be.
Should I tell a customer why their price is going up? Briefly, honestly and without over-explaining. Cost, scope or the fact that it has not changed in several years. Long justifications invite negotiation.
What if they will not sign an agreement? Then you have learned something worth knowing. A customer unwilling to document what you have already been doing is telling you how a dispute would go. Weight that in the risk column.
Is it ever right to fire a profitable customer? Yes. When they damage staff retention, when the risk cannot be contained or when they conflict with what you are building. Just make it a priced decision rather than a reaction to a bad week.
The audit is usually straightforward. The sequence afterward is where engagements concentrate, because that is where the relationship is either kept or damaged. If you want help running it on your actual accounts, here is how I work.