9 Hidden Operating Frictions Reducing Customer Profitability
Every function is well run. The profit still leaks, because it leaks at the joins between them — and nobody reports on a join.
- Customer profitability is lost between functions rather than inside them. Each department is run well and the total is worse than the parts.
- The nine frictions below all sit at handovers — where work passes from one team to another and nobody owns what happens in between.
- The position worth holding is not “every function is performing”. It is “we know what a customer costs to serve across all of them, and someone owns that number”.
In most mid-market firms, every function is genuinely well run. Sales performs. Operations delivers. Service responds. Finance reports accurately. And customer profitability is lower than any of those performances would suggest.
That is not a contradiction. Profit leaks at the joins between functions, not inside them. Each department optimises what it is measured on, which is the right thing to do given how it is measured. What happens in the space between departments belongs to nobody.
Those spaces are invisible in the reporting, because reporting follows the organisational chart. Every function reports its own numbers. Nothing reports on a handover.
The nine frictions below all live in that space. None of them is anyone’s fault, and none can be fixed by the function where the symptom appears.
What sales promised is news to the people delivering it
Deals get won on specifics. A timescale, a bespoke arrangement, a service level, a small commitment made to get the signature. Each one is reasonable in the room.
Those commitments frequently do not travel with the account. Operations discovers them when the customer expects something the standard process does not provide. Service discovers them during a complaint.
So the cost of honouring them lands somewhere nobody planned, and the margin on that account is lower than the deal sheet said. The fix is a structured handover that captures every commitment, and a rule that anything non-standard gets costed before it is agreed rather than after.
The customer explains themselves repeatedly, and each retelling costs money
Each function holds its own record. Sales in the CRM, service in the ticketing system, operations in the delivery platform, finance in the ledger. Each is accurate for its purpose.
The customer meets all of them and none of them knows what the others were told. So they repeat the context each time, and each repetition costs a colleague’s time as well as a little goodwill.
The cost is spread thinly enough to be invisible and large enough to matter. The fix is one shared view of the customer that every function can see, which is a governance decision far more than a technology purchase.
Exceptions get handled brilliantly and never get counted
Something goes wrong and people sort it out. A manager intervenes, a workaround gets applied, someone stays late. The customer is looked after and the issue closes.
What almost never happens is anyone recording what the exception cost or why it occurred. So the same exception recurs, gets handled equally well, and consumes the same effort again.
Firms with excellent people are especially exposed to this, because good people absorb problems rather than escalating them. The fix is to log exceptions with a cause and a cost, then treat the top three as work to be designed out rather than absorbed.
Reporting follows the organisation chart. Profit leaks at the handovers, and nothing in the pack reports on a handover.
Pricing is set centrally and eroded locally
The pricing framework is agreed. Discount authority is delegated so deals can move. Each individual concession sits within someone’s limit and is defensible on its own.
What is rarely visible is the aggregate. Nobody reports realised price against list across the whole base, by segment, by salesperson, over time. So the erosion accumulates quietly and is discovered during an annual review, by which point it has been repriced into customer expectations.
The fix is monthly reporting on realised price rather than annual, with the pattern visible by person and segment. It changes behaviour simply by being seen.
Service level commitments were agreed without operations in the room
Response times, delivery windows, availability guarantees. These get agreed commercially, because they win business, and they are usually achievable.
What is often not modelled is what achieving them consistently costs — the standby capacity, the expedited delivery, the overtime when volume spikes. Those costs land in operations, where they appear as an operational overrun rather than as the price of a commercial promise.
So operations gets challenged on cost control for delivering exactly what sales sold. The fix is to cost service commitments before they are offered, and to report the cost against the account that carries them.
Knowing what a customer costs to serve across every function is the first step.
A free Executive AI Opportunity Assessment. Forty-five minutes, online, with your leadership team. We arrive with a model of how firms in your sector make money and where AI has moved those numbers — and you tell us where it is wrong.
Book a free assessmentBad data enters at one end and costs money at the other
An address is wrong. A product code is mistyped. A contact is out of date. Each error takes seconds to make and is usually nobody’s fault in particular.
Downstream it produces a failed delivery, an incorrect invoice, a chased payment, a service call. Each of those costs many multiples of what correcting the data at entry would have cost.
The team that creates the error never sees the consequence, and the team that pays for it cannot trace it back. The fix is validation at the point of entry and a feedback loop that shows the originating team what their errors cost downstream.
Each department performs against what it is measured on. The space between them belongs to nobody.
Every function reports its own numbers. Nothing reports on a handover.
Good teams fix exceptions rather than escalating them, which hides the cause and repeats the cost.
By which point the erosion has been repriced into what customers expect.
This describes a pattern rather than a measured dataset. The test worth running is your own: take your ten largest customers and work out what each actually costs to serve across every function. Most firms find the ranking is not what they expected.
Nobody knows what an individual customer costs to serve
Revenue by customer is known precisely. Cost to serve by customer usually is not, because the costs sit across functions and none of them allocates by account.
So decisions get made on revenue. The largest customers get the most attention, the best terms and the most flexibility — which is right if they are also the most profitable and expensive if they are not.
The fix is a cost-to-serve model by account or segment, built once and refreshed periodically. It is a few weeks of finance and operations work, and the ranking it produces is usually a surprise.
Three questions worth asking this week
- Take your ten largest customers by revenue. Can anyone rank them by profit? If not, the biggest accounts are being managed on the wrong number.
- What were the top three causes of exceptions last month, and what did each cost? If nobody logs them, they are being absorbed rather than fixed.
- What is realised price against list, by segment, over the last twelve months? If it is only reviewed annually, the erosion has already happened.
Functions are measured on things that quietly conflict
Sales is measured on revenue. Operations on cost and efficiency. Service on resolution time. Finance on working capital. Each measure is sensible.
Together they pull against each other. Sales wins volume that operations has to absorb inefficiently. Operations optimises batches that delay what service promised. Service resolves quickly by giving things away that finance then reports as margin loss.
Nobody is behaving badly. Everyone is doing what they are measured on. The fix is not better cooperation — it is a shared measure that at least two functions carry together, so the conflict has somewhere to be resolved other than in a meeting.
Nobody owns the customer across the whole journey
The most structural of the nine. Sales owns winning. Operations owns delivering. Service owns problems. Finance owns billing and reporting. Every function has an owner.
What does not exist in most mid-market firms is anyone who owns the customer across all of it — with authority over pricing, service commitments, delivery arrangements and exception policy, and accountability for what each account actually returns.
That absence is what the other eight describe. Each is a decision made well inside one function with a consequence that lands in another. The fix is to make the end-to-end customer someone’s explicit responsibility at executive level, with authority across the functions rather than inside one of them.
What this means
None of these nine are performance failures. Every function is doing what it is measured on, competently. That is precisely why the frictions survive — nothing in the reporting shows them, and nobody is accountable for the space they live in.
The pattern is consistent. Decisions get made locally and well. Their consequences land elsewhere, where they appear as somebody else’s cost problem. The originating function never sees what it caused and the receiving function cannot trace it back.
The economics of fixing it are unusually good, because almost all of the work is analysis and measurement rather than build. Costing customers to serve takes weeks. Logging exceptions takes a form. Reporting realised price monthly takes a change of frequency. None of it needs a platform.
Which is the same argument we make about AI. Change the work before you change the technology — and before automating a process that spans four functions, it is worth knowing which of the joins is currently costing you money.
Can you rank your largest customers by profit, or only by revenue?
Forty-five minutes with your leadership team, online, at no cost. We bring a working model of how firms in your sector make money and where AI has moved those numbers. You tell us where it is wrong. Then we tell you what we think — including “don’t invest”, if that is the honest answer.
Book a free assessment45 minutes · online · no cost · nothing to prepare
Continue reading
9 Hidden Lifecycle Frictions Quietly Raising Your Acquisition Cost
The same logic applied to winning customers rather than serving them.
For retail11 Hidden Costs Quietly Undermining Retail Transformation
Where complexity grows faster than the discipline meant to contain it.
For software firms8 Fragmentation Patterns Limiting Mid-Market Software Growth
Revenue scales faster than the way the business is organised around customers.