Every explanation of why sales teams overstuff the first deal lands in the same place. They're paid on the close and on nothing after it. The consequence shows up three quarters later, in a number nobody who touched the deal is measured on. Fix the comp plan and you fix the behaviour.
I believed that too, until I watched it fail.
The company that paid on retention anyway
A company I worked with had its reps on recurring commission. Not a spiff, not a one-year accelerator. Real ongoing money for as long as the customer stayed.
They front-loaded anyway. Discounts to get it signed. Modules thrown in that the buyer couldn't use yet. And on the calls, the line that quietly sets a date for the customer to decide: give it three months and see where it lands.
By any model of incentive design that shouldn't happen. They were being paid to care about month fourteen.
So I asked them, and the answer was better than the theory.
They didn't think the customers would be there at month fourteen. They'd watched delivery up close. They knew what onboarding actually looked like and what happened after it. The recurring commission wasn't income to them. It was a lottery ticket on a company they'd stopped believing in.
And if the backend is worth roughly nothing, the rational move is to make the front end as big as you can.
They were right, which is the problem
This is the part that took me a while.
Their forecast was accurate. Customers were cancelling at around three months. The reps weren't being cynical, they were being observant, and they were better informed than anyone above them, because they were the ones making the promises and then watching what happened to them.
But the way they responded to an accurate forecast made it more accurate.
The discount set a number that account would be measured against for the rest of its life. The thrown-in modules sat there doing nothing, which is worth nothing and reads as evidence the core isn't working. And give it three months told the buyer exactly when to decide.
Every one of those shortens a lifetime. So the reps predicted short lifetimes and then structured deals that produced them. The forecast and the cause were the same act.
Deal structure is a forecast
Here's what I take from it, and it's more useful than the comp lesson.
Your sales team is the earliest well-informed read on your retention that exists inside your company, and they will never tell you in a meeting. They report it through how they build deals.
A rep who believes the customer will still be there next year has a reason to leave something on the table, because it's worth more later and they'll get paid on it either way. A rep who doesn't takes everything now.
So the diagnostic is simple. If your comp already rewards retention and your reps are still front-loading, you don't have a comp problem. You have a delivery problem your sales team has already priced in.
That reframes what a lot of leaders read as a discipline issue. Overstuffing looks like greed. Under aligned comp, it's closer to an unusually honest forecast.
Which means the fix runs in a specific order
Ask a rep to hold something back for a milestone six months out and you are asking them to believe the customer will be there in six months. If they don't believe it, you're asking them to set money on fire for the good of a company they've watched miss.
No comp plan survives that. You can pay people to care about month fourteen and still lose, because caring about month fourteen requires believing in month fourteen.
So it isn't comp, then behaviour. It's delivery, then belief, then behaviour. Which is the same order every expansion argument runs in, arriving from a direction that one doesn't cover.
What it looks like when they believe you
The version that works doesn't require convincing anybody of anything.
When a rep is confident the customer is going to be around, holding something back stops being a sacrifice and starts being obvious. The item prices higher at the milestone that earns it than it ever did inside the bundle. The rep gets paid on it. The customer gets it when they can actually use it, which is the only time it was ever worth anything.
Nobody has to be talked into that. It's just arithmetic, and it only works if the first part is real.
Your reps have already done that arithmetic. Go and read what they built.
Lincoln Murphy formally named and popularized Customer Success starting in 2010 and has spent 15 years connecting it to expansion revenue and commercial outcomes. Read The Premise.