Listen to how a customer describes what they bought.
"I upgraded." "We added capacity." "I bought the next tier." Those are active. The customer is the subject of their own sentence, and the sentence is neutral or good.
Now the other one.
"I was upsold."
Nobody has ever said that happily. Not once, in the history of the phrase. It's passive, it's always a complaint, and the customer has told you exactly what happened to them. Something was done. They were the object.
Same transaction. Same money on the same invoice. Two completely different grammars, and the grammar is the diagnosis.
The sale works. That's the problem.
An upsell doesn't fail at the moment of sale. It closes. The money moves, the number lands, somebody gets credit, and the account shows up on a report as expanded.
The regret arrives a few days later.
And here's the signal almost nobody recognizes: a lot of those customers ask to reverse it. They come back, slightly embarrassed, and ask whether they can undo the thing they just agreed to.
That question is the most useful sentence the account will ever produce, because it's a customer telling you in writing, at no cost to you, that the sale was wrong. You will not get a clearer one all year.
And they get talked out of it.
Why you talk them out of it
Not greed. It's rarely greed, and reading it as greed means you'll fix the wrong thing.
It's scarcity, and it's structural rather than personal.
Expansion in that company is haphazard. Nobody owns it, nothing triggers it, no forecast contains it. So when one arrives, it arrives looking like luck, and you don't give luck back. That revenue is rare, it's already booked, and handing it over feels like setting money on fire.
Now picture the company that can produce expansion on purpose. Named accounts, observable gates, a number that was measured rather than invented. That company gives the money back without blinking, because another one is already in the forecast for next month.
The absence of a motion is what makes the accidental sale impossible to release. You're not clinging to the revenue. You're clinging to the only one you've got.
The waiting period reads as health
So the reversal gets declined, warmly, with reasons. And the customer stops asking.
That's the part that costs you. The asking was the alarm, you answered it, and the problem it was attached to went right on running underneath a system that now has nothing left to report. No ticket, no complaint, no angry call to anybody senior.
What replaces it is a customer sitting on a decision they've already made, waiting for the first moment they have any real leverage over you. That moment is renewal, and it's eleven months away.
They show up with usage data, a line-item list, and twelve months of quiet resentment, and they do one of two things. They right-size the account down to what they actually use. Or they leave entirely.
Neither of those is a renewal-cycle event. The decision was made at purchase and everything since has been a holding pattern that every dashboard you own reported as a healthy expanded account.
They've been planning their exit since you upsold them.
What to do with this on Monday
Two things, and the first one takes ten minutes.
Go read what your customers actually wrote. Search your inbox and your ticket queue for "upsold." Then search for "upgraded," "added," "moved up to." Look at which accounts use which voice, and then look at what happened to those accounts at renewal. The pattern will be in there and you didn't have to instrument anything to see it.
Second: find out what happens at your company when a customer asks to reverse an expansion. Ask the person who'd take that call. If the answer is that they talk them through it and keep the revenue, you have a policy that converts your best available signal into twelve months of silence.
The fix isn't a better rebuttal. It's selling the thing at the point the customer arrives at it, so nobody has a reason to ask.
An upsell is complete when the money moves. The bill arrives a year later, and by then nobody remembers ordering it.
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.