Upsell Call Teardown: The Check-In That Turned Into a Downgrade Request
A real-shaped expansion call annotated line by line: the throwaway opener, the premature feature pitch, the contradicted metric, and the missed stakeholder — then the same call done right.
I have listened to a lot of expansion calls. The ones that go badly rarely go badly because the rep was rude, unprepared, or unlikeable. They go badly because of a sequence. Four small decisions, each one defensible on its own, that stack up until the customer is the one running the call and the topic has quietly changed from "more" to "less."
What follows is a teardown of one of those calls. It is a composite — the shape is real, the specifics are changed, and no customer or employer is being named here. But if you have ever booked a friendly quarterly check-in and hung up wondering how you ended up defending the original purchase, you will recognize every line.
Set the scene. Mid-market SaaS, a workflow tool, roughly forty seats on the account. Renewal is a quarter out. The AE has a standing check-in on the calendar and a number to hit, and the plan for this call is seats plus a tier upgrade. The customer contact is an ops manager who signed the original deal and has been perfectly pleasant for a year.
Here is how it went.
The call
Minute one
Rep: "Hey Dana, thanks for hopping on. Nothing urgent, just wanted to check in and see how everything's going."
Customer: "Sure, yeah. Things are fine. Actually, while I have you — do you have a usage report? Finance is asking me for one."
That is the whole game, decided in two lines. The rep opened by telling the customer the call had no agenda and no stakes. The customer, being a reasonable adult with a full calendar, immediately supplied one of her own. And her agenda is a finance-driven audit of what this tool is costing.
Notice she did not ask a hostile question. "Do you have a usage report" is neutral on its face. But the reason behind it is not neutral, and the rep never found out what the reason was.
Minutes two through six
Rep: "I can pull that for you, absolutely. Before I do — I wanted to walk you through the new automation module we launched. It's pretty exciting. It lets you build multi-step approvals without touching the admin panel, and a lot of teams are moving to it."
Customer: "Okay. Is that included?"
Rep: "It's on the Growth tier, so it'd be an upgrade from where you are today. Happy to put pricing together."
Customer: "Got it. Let's come back to that. Can you send the usage report?"
The rep pitched a feature before establishing a single unit of value the customer had already received. So the customer priced it against nothing. When you introduce a paid upgrade before you have re-anchored what the current spend has produced, the customer does the only math available to her: current bill, plus more, in exchange for a capability she did not ask about.
Also note "Let's come back to that." Nobody ever comes back to that.
Minutes seven through fourteen
Rep: "So on usage — I pulled some numbers before the call. You've got forty seats provisioned and we're seeing about twenty-eight of them active in the last thirty days."
Customer: "That's lower than I thought. My team lead told me almost everyone's in there daily."
Rep: "Well, our definition of active is a login plus at least one action. Some of your users might be getting value passively, from reports and notifications, so the real number is probably higher than what the dashboard shows."
Customer: "Hm. So the number in the report you're about to send me is wrong?"
This is where the call actually died, and it died on a line the rep thought was helpful.
The customer offered her own belief — "my team lead told me almost everyone's in there daily" — and the rep contradicted it, then contradicted his own data to soften the contradiction. Now there are three conflicting versions of reality in the room: the dashboard, the team lead, and the rep's caveat. The customer cannot defend a renewal internally with three versions. She can only defend a smaller number, because a smaller number is the safe one.
When a customer hands you a metric they believe, you do not correct it in the same breath. You get curious about the gap. The gap is the entire conversation.
Minutes fifteen through nineteen
Customer: "I should mention, we've got a new VP of Operations starting in a few weeks. She's going to want to look at all of this anyway."
Rep: "Oh, congrats! Yeah, happy to do an onboarding session with her once she's settled. I'll follow up."
Customer: "Sure. And on the seats — is there flexibility if we needed to come down at renewal? Not saying we will. Just want to know the options."
Rep: "There's usually some flexibility, yeah. I'd have to check the terms. But before we go there, I really do think the automation module would help with the adoption piece —"
Customer: "Let's start with the report."
A new VP of Operations was mentioned out loud, unprompted, and the rep treated it as social news. That was the exit ramp. That was the moment to stop the check-in entirely and start a different conversation, and the rep drove past it at speed to get back to the automation module.
Then he answered the downgrade question. Honestly, helpfully, and fatally. "There's usually some flexibility" is the sentence the customer will repeat to finance, and she will repeat it without the hedge.
Call ends. Rep logs it as positive. Rep sends the usage report. Ninety days later the account renews at fewer seats and the rep tells his manager the customer had budget pressure.
The customer had budget pressure before the call. What the call did was hand her the evidence.
The four turning points
1. The throwaway opener framed the call as optional
"Just checking in" is not a greeting. It is a statement about the value of the next twenty minutes, and the value you assigned is zero. Every customer hears it as permission to redirect.
The fix is not to be aggressive. It is to bring a reason. A specific, customer-side reason that you state in the first fifteen seconds, before they can supply their own. "I pulled your last quarter before this call and there's one pattern in it I want your read on" does the same social work as "just checking in" while establishing that you did homework and that the twenty minutes have a subject.
If you sell into industries where the check-in is a cultural ritual, this matters more, not less. The real estate version of this call is almost entirely made of people who picked up expecting nothing and will happily give you nothing back if that is what you ask for.
2. The feature pitch landed before any value recap
Order of operations. Value first, always, and value in the customer's own units, not yours.
The rep had a year of account history available and led with a module launch. What he should have led with is what the account produced: cycle time on approvals, requests processed, hours the ops team did not spend chasing signatures. Whatever this product actually does, in numbers Dana would use in her own status update.
Once the customer has agreed out loud that something worked, the upgrade is an extension of a working thing. Before that, it is an increase to a line item.
This is the single most common ordering mistake I see, and it is the reason the upsell script for SaaS customers is built around surviving a seat utilization report rather than avoiding one. The report is coming. You want to be the person who brought it.
3. The rep contradicted the customer's own metric
There are two ways to lose an argument with a customer about data. You can be wrong, or you can be right. Both cost you the account.
When Dana said her team lead reported daily usage, the correct move was not correction. It was: "That's interesting, because the dashboard is showing something different, and I'd rather figure out which one is real than send you a report you can't defend. Who would know?"
That sentence does four things at once. It validates her source. It flags the discrepancy without assigning blame. It positions the report as something she has to defend internally — which she does — and it makes you the person helping her defend it. And it produces a name.
Instead the rep undermined his own dashboard to be nice. Never do that. If your data has caveats, state them before you state the number, not after the customer reacts to it.
4. The new stakeholder was mentioned and skipped
A new VP of Operations arriving before renewal is the largest single piece of information in this call, and it was delivered for free.
New executives audit. That is the job. She will arrive, ask what every tool costs, ask who owns it, and cut the ones nobody can defend in one sentence. Dana knows this, which is exactly why she mentioned it, and why she asked about seat flexibility ninety seconds later. Those two lines are the same line.
The exit ramp was: stop selling the module, and get in front of the VP before the audit rather than after it. "When does she start? Here's what I'd suggest — let me build you a one-page summary of what this has done for your team over the last year, in your numbers, so that when she asks you're not building it from scratch at eight in the morning. And if it's useful, I'll join that conversation."
Now you are not a vendor being evaluated. You are the person who made Dana's first meeting with her new boss easier. That is a different relationship, and it survives audits.
The same dynamic shows up anywhere renewals are seasonal and contacts rotate. In staffing and recruiting, the quarterly check-in and the redeployment conversation are the same call, and missing a new hiring manager on the account costs you the whole quarter.
The rewrite
Same account. Same forty seats. Same usage gap. Same VP arriving. Here is the call done in the right order.
Rep: "Dana, thanks for the time. I went through your last quarter before this call and there's one thing in there I want your read on, plus I've got a usage summary for you — I'm guessing finance has been asking."
Customer: "Ha. They have, actually."
You brought the thing she was going to ask for. The audit is now happening on your terms.
Rep: "Figured. Before I hand it over, quick recap so the numbers have context. Last quarter your team ran approvals through here that used to sit in email — your side of the process went from days to same-day on most of them. Does that match what you're seeing?"
Customer: "Yeah, that's been the big one. The approvals thing."
She said it out loud. Write it down verbatim, because that is the sentence that goes in front of the VP.
Rep: "Good. Now the part I want your read on. Forty seats provisioned, twenty-eight active by our definition — login plus an action in thirty days. Before you take that anywhere, I want to know if it's real."
Customer: "That's low. My team lead says almost everyone's in there daily."
Rep: "That's exactly the gap I wanted to ask about. One of those two things is wrong and I'd rather we find out now than have you present a number you can't back up. Can we get your team lead on for fifteen minutes this week? If those twelve people are working through someone else's login or living in the notifications, that's fixable and the report changes. If they genuinely aren't using it, I'd rather know that too."
Customer: "...Yeah, okay. Let me set that up."
Notice what did not happen. No defending. No caveat softening the number. No upgrade pitch. The rep turned a threatening metric into a joint investigation with a next meeting attached, and picked up a second stakeholder in the process.
Customer: "I should mention we've got a new VP of Operations starting in a few weeks."
Rep: "When does she start? — Okay. Then here's what I'd suggest. She's going to ask you what every tool costs in her first month, because that's what new VPs do. Let me build you a one-page summary: what the approvals change was worth in your numbers, where the twelve seats actually stand once we've talked to your team lead, and what the options are going into renewal. You walk into that meeting with it already done. If she wants to talk to me directly, I'm available."
Customer: "That would genuinely help. She's going to want all of that."
Rep: "One more thing and then I'll let you go. When she asks what's next, what's the answer? Is there a part of the process that's still living in email?"
Customer: "Honestly? Multi-step approvals. Anything with two or three sign-offs still goes around us."
Rep: "Then let's put that in the summary too, because that's the thing we shipped last quarter and it's on a different tier. I'm not pitching it today. I'd rather it show up as an option in a document you built for your new boss than as a vendor upsell in the same month she starts."
The module got pitched. It just got pitched after the value recap, after the metric was jointly owned, and after it had been reframed as Dana's idea in Dana's document. That is the entire difference.
What the customer says instead
In the first version, the last thing the customer says is "is there flexibility if we needed to come down." In the rewrite, the last thing she says is a request for help preparing for her new boss, with two follow-up meetings booked and a feature gap she named herself.
Same account. Same data. Same quarter. The number of seats did not change during the call — the story about them did.
If you have a QBR next week, the practical version of all this is short. Bring a reason in the first fifteen seconds. Recap value in their units before you name a product. When their metric conflicts with yours, investigate instead of correcting. And when a new name enters the conversation, stop everything else and go get that meeting. The same four beats hold whether you are selling software, freight and 3PL services where the check-in call is the entire account management motion, or anything else with a renewal date on it.
Knowing this and doing it under pressure are different skills, and the pressure in an expansion call is real — it is your customer, your renewal, and your number. What I would do before the QBR is run the hard version of the call out loud a few times, with someone playing a Dana who asks for the usage report in the first thirty seconds and mentions a new VP halfway through. That is what we built DrillCall for: rehearsing the specific moment you know you will fumble, until the response is boring. The call you want to be good at is not the friendly one. It is the one where the customer opens with a question you did not plan for.
Go run it before Tuesday.