The commission goes up. For a few weeks the floor sounds different — people stay on the phone longer, the pipeline moves, the manager who signed off on the raise feels vindicated. Then it settles, and the numbers come back to where they were, with a higher cost per sale attached to them.
That pattern is worth taking seriously rather than reading as ingratitude. A raise changes what a job pays. It does not change what the job is. If what is draining the team is the shape of the day — the volume of refusals, the hours lost to work nobody needed a salesperson for, a target nobody on the floor can move — then more money is a painkiller with a known half-life.
Why a raise wears off
Pay is a floor, not a fuel. Below a certain number people leave; above it, more money stops buying more effort fairly quickly, because effort is not the thing that went missing. What went missing is usually one of three: the sense that the work is going somewhere, the sense that the result depends on you, or a day that is not mostly grinding.
It is also worth looking at who a bigger commission actually pays. Strong sellers were already motivated and now cost more. Struggling ones do not improve because the prize grew — if somebody cannot open a conversation, doubling the payout for closing one does not teach them how. The raise rewards the people who did not need it and skips the people who did.
A raise changes what the job pays. It does not change what the job is.
What a motivation system is actually made of
It helps to separate the parts, because in practice only two of them ever get adjusted.
- Base pay. The part that says the job is a job: predictable, enough to live on, not a bet. Its purpose is to take financial anxiety off the floor, not to reward performance.
- Variable pay. The part tied to results. Its purpose is direction — it tells the team which of the many things they could do this month the company actually wants done.
- The conditions of the day. Which hours go to selling and which go to data entry, how leads arrive, how much rejection the role absorbs, what tools handle the boring parts. This is the part people are describing when they say they are tired.
- Recognition and a path. Whether good work is seen by name, and whether there is somewhere better to be later on that is worth working toward. Cheap to provide, easy to forget.
The first two are financial, and they are the two you can change in an afternoon with a spreadsheet. The other two require changing how the work is organised, which is slower and far less satisfying to announce. So the lever that gets pulled tends to be the one that was not broken.
The KPIs that demotivate
A metric is a promise about what counts. Pay on the wrong one and you have not merely failed to motivate — you have taught the team something you did not mean to teach them.
- Targets nobody on the floor can move. A revenue number that turns on the marketing budget, the season or the product roadmap. A target you cannot influence is not a target, it is a weather report.
- Call volume quotas. Pay for dials and you get dials: short ones, careless ones, to whoever sits nearest the top of the list. The activity rises and the pipeline does not.
- Minimum call duration. This one is worse, because it turns your best sellers into your slowest. Somebody who qualifies a lead quickly is penalised for doing it well, so they learn to pad.
- Another department's results. A bonus that hinges on delivery, support or finance hitting their own numbers. The seller is now accountable for people they cannot instruct.
- A supervisor's opinion, scored on a scale. Subjective scoring with no written standard reads as favouritism whether or not it is, and it cannot be argued with — which is the part that corrodes.
What these share is not difficulty. It is that the link between what a person does and what a person gets has been cut. Once that link goes, the rational move is to stop trying to shift the number and start managing how it looks — and that habit, once learned, does not stay politely inside the metric that taught it.
When money no longer works, audit the day
Before touching the compensation plan again, spend a week finding out what the job actually consists of. Four things, in this order.
How much of a seller's day is spent talking to somebody who might buy? The rest — dialling numbers that do not answer, retyping what the CRM already knows, chasing lists, first-touch calls to people who will never qualify — is the real subject of the conversation. If your team is arguing about pay, this is the thing worth checking before you concede the argument.
Not a leaderboard for its own sake: one definition of a qualified lead, one place the pipeline lives, one set of figures anybody can check. Transparency motivates mainly because it makes fairness verifiable.
If the strongest leads go to whoever asks loudest, or to whoever the manager likes, no incentive scheme survives it. The rule can be round-robin, by territory, by scored quality — what matters most is that it is written down and applied the same way on a bad week.
Whatever is left that needs no judgement — the first touch, the reminder, the confirmation, the follow-up nobody got to — is work for a system, not for somebody you hired to close.
Routine is the thing doing the damage
Cold calling earns its own paragraph, because this is where the damage concentrates. A day of first-touch calling is mostly a day of being told no by strangers, at a pace that leaves no room to recover between them. That is not work people build a tolerance for. They build an aversion to the phone, and then a CV.
The uncomfortable part is that the work does need doing. Lists have to be called, interest has to be found, and the person who eventually has the useful conversation has to arrive at it somehow. The open question is only who does the first, mechanical half.
Which is the argument for handing the first touch to an AI voice agent and keeping experienced people for the conversations where experience pays: the qualified prospect, the real objection, the deal that needs a judgement call. The agent is not worn down by a day of refusals and is not owed a raise for absorbing them. Your seller opens the day with a list of people who have already said they are interested — which is the job they thought they were taking.
Where these figures come from
Ten years of running call centres, up to fifteen hundred live operators at peak, across France, Mexico and beyond. What this article prescribes comes out of that history and out of our own operating floors: it is not research and it is not an industry standard. One of those floors is worth naming, because it decides how much of the above a manager can even see. Under manual quality control, under 5% of recorded calls are ever reviewed by anybody, which is why “we listen to our calls” and “we know what happens on our calls” are two different sentences. Our own scoring runs over every recorded call instead of a sample, which is what makes a fairness rule checkable rather than merely stated.
None of this is an argument against paying people well. It is an argument that pay is the floor of a motivation system rather than the whole of it, and that when a raise stops working the answer is usually somewhere in the day rather than in the plan. To see what moving the mechanical part of the calling off your team would cost against an hour of a seller's time, the calculator does that arithmetic with your own numbers.