How to Turn Usage Data Into an Expansion Case Without Sounding Like an Auditor

The stat that ends the call before it starts

Here is the opening I hear more than any other on expansion calls, and it is almost always delivered with pride:

"So before we get into the roadmap stuff, I pulled your usage report. You're licensed for a lot more seats than are actually logging in, and I wanted to walk through that with you."

The rep thinks they have just demonstrated diligence. What the customer heard was: my vendor has been going through my drawers.

And the second thing the customer heard, maybe half a second later, was: this is the setup for a price increase. Because in the customer's experience, nobody opens with a number about your account unless the number is about to cost you something. Utilities do this. Insurers do this. Their own finance team does this. The pattern is so well worn that the reflex fires before the customer has even processed what you said.

So the rest of the call is spent with the buyer's hands up. Every follow-up question you ask sounds like discovery for a bill. Every silence you leave gets filled with a defense of why the numbers look the way they look. You wanted a diagnosis and you got a deposition, and you were the one holding the clipboard.

I have sold inside two very large companies, and both of them had beautiful adoption dashboards. The dashboard is not the problem. The problem is that a dashboard is evidence, and evidence has a genre. If you present a number the way a prosecutor presents a number, the customer will play the part the genre assigns them.

Why the number reads as a bill

There are three specific things that make usage data feel like an audit, and once you can name them you can strip them out.

The first is that the customer didn't ask for it. Unsolicited measurement of someone else's behavior is, socially, an accusation. It doesn't matter how neutral the number is. If I tell you unprompted how many times you went to the gym last month, you are going to hear a judgment in it, because why else would I have counted.

The second is that the number is about the whole account. Account-level metrics are the language of contracts and renewals. When you say "your organization is at X," you have addressed the customer as a line item rather than as a person with a team and a set of problems. Nobody feels like an organization. They feel like the person responsible for one piece of it.

The third, and this is the one most reps miss, is that the number is presented as a finding rather than as a puzzle. A finding has a conclusion baked in. A puzzle invites the other person to think. "Twelve of your forty seats are dormant" is a finding. "There's a pattern in here I can't explain, and I was hoping you could" is a puzzle. Same data. Completely different room.

The fix is not to hide the data. The fix is to change what the data is for. Usage numbers are a diagnostic instrument, like a stethoscope. You do not open with the stethoscope. You open with "tell me where it hurts," and the instrument confirms or complicates what they told you.

Three numbers, and only three

Most adoption dashboards give you dozens of fields, and the temptation is to bring all of them because they were expensive to build. Bring three. More than three and you are presenting, and presenting is the audit posture.

One: where the product is working best

Not the average. The peak. Which team, branch, pod, region, or individual is using the thing most completely? What are they doing with it that nobody else is?

This is the single most useful number you can carry into an expansion conversation and it is the one reps almost never lead with, because it doesn't create urgency. That is exactly why it works. It reframes you from someone counting unused licenses to someone who noticed something good.

In a staffing account, that might be the one recruiting pod that actually re-engages placed candidates instead of letting them go dark. In a freight account, it's the lane team that logs exceptions in the system instead of over the phone. In a SaaS account, it's the department that has built its weekly ritual around your reporting view rather than exporting to a spreadsheet.

You want to be able to describe that group's behavior in a sentence a human would say out loud, not in a metric. "Your Midwest desk runs their Monday standup inside the tool" is worth more than any engagement score.

Two: the spread between the best group and everyone else

This is the number that does the actual work, and it is a comparison inside the customer's own house, which is why it doesn't feel like an attack from outside. You are not telling them they are underusing your product relative to your other clients. You are telling them one part of their business has figured out something the rest hasn't.

The spread is powerful because it pre-answers the objection you would otherwise get, which is "our people just don't work that way." Clearly some of their people do. The question stops being whether the behavior is possible in their culture and becomes why it hasn't traveled.

Be careful with how you name the low end. Never name a low-performing manager, ever, even if the customer names them first. You are not there to hand anyone ammunition for an internal fight. Say "the other regions" or "everyone outside that group." If your champion wants to be specific, let them be specific. That is their call to make, not yours.

Three: where people are pressing against the edge of what they bought

This is the expansion signal, and it is behavioral, not numeric. Where are users doing something awkward because the thing they actually need sits outside their current package?

Exports to spreadsheets. Duplicate records created to work around a permission model. Shared logins. A support ticket asking whether a feature exists in a module they don't own. Manual re-keying between your system and something else. Requests to your CSM for a report you can't build in their tier.

Those behaviors are the customer telling you where the next budget lives, in their own handwriting, without being asked. And unlike a dormancy number, they are flattering. Nobody is embarrassed to be caught working hard around a limitation. Plenty of people are embarrassed to be caught not logging in.

Three numbers. Peak, spread, edge. Everything else stays in the tab you don't share.

The framing sentence

There is one sentence that separates an expansion conversation from an audit, and it goes roughly like this:

"I'm not here to tell you how you're using it. I want to show you what your best team is doing that the others aren't, because I think there's something in it and I can't tell from the outside whether it's the tool or the people."

Read that again and notice what it does. It disclaims the audit explicitly, which matters, because the customer's reflex is already firing and you have about one sentence to interrupt it. It centers on a group that is succeeding, which changes the emotional temperature of the whole call. And it ends by admitting you don't know the answer, which is the part that makes it a conversation instead of a presentation.

That last clause is doing more than it looks like. "I can't tell from the outside" is a genuine statement of your position. You have telemetry. You do not have context. You do not know that the Southeast branch is understaffed, that the ops lead there is on parental leave, that the whole region is being restructured. The customer knows all of that. Saying so out loud hands them the expert role, and people defend themselves much less when they have been cast as the expert.

Then stop talking. Really stop. The instinct after a setup that good is to keep going into the data, and if you do, you have converted your own framing back into a presentation.

Let them draw the conclusion

The whole point of leading with the peak and the spread is that the obvious question forms in the customer's head without you asking it. Why is that group different, and what would it take to get the others there?

When the customer says that out loud, you have won the call, and the deal that follows is almost administrative. When you say it out loud, you are selling, and everything after it gets discounted as sales talk.

So build in the pause. After you describe what the best group does, ask something narrow and genuinely curious rather than something leading. "Do you know how that started over there?" is good. "Was that deliberate or did it just happen?" is good. "Wouldn't it be great if everyone worked like that?" is not — that is a rhetorical question, and rhetorical questions are the audit's cousin.

Half the time the answer surprises you. The best group is best because of a manager who used to work at a company that ran the same software, or because they inherited a template from someone who left. That origin story is the actual expansion path, and you cannot guess it from a dashboard.

The mechanics of holding that pause, and what to say when the customer fills it with a defense instead of a question, is the meat of the SaaS upsell script that survives a seat utilization report, because seat data is the version of this problem where the audit reflex is strongest.

The question that finds the next budget

Once the customer is diagnosing rather than defending, there is one question I would not run an expansion call without:

"If that team's way of working became the standard everywhere, what would break first?"

It is a question about constraints, and constraints are where money is. The answer is almost always one of a few things: we don't have the headcount to support it, the data wouldn't hold up at that volume, compliance would need to sign off, or we'd need something we don't have today.

Every one of those is a scoped expansion conversation, and the customer just scoped it for you. You did not propose a module. They described a gap, and the module is the answer to the gap.

The follow-up matters as much as the question. "Whose problem would that be?" Not to be political, but because you need to know whether your champion can fund the fix or whether this becomes a multi-threaded deal. In a regulated account the answer is frequently a name your champion does not control, which is why the insurance version of this conversation spends so much time on getting from a claims account review to the person who owns the recovery budget.

And if the answer is "nothing would break, we just haven't gotten around to it" — believe them. That is not an expansion. That is an enablement problem, and if you sell them more of something they haven't rolled out, you have bought yourself a churn event with a delay on it.

Never run the expansion on a complaint call

This is the rule I would enforce hardest if I ran your team, and it is the one most often broken by good reps with quota pressure.

If the customer has raised a complaint on this call — an outage, a bad support experience, a bug that cost them a day, an invoice error — the expansion conversation is dead for the duration of that call. Not postponed until later in the call. Dead. You do not pivot after resolving it. You do not say "and while I have you."

The reason is simple. A complaint puts you in debt. Selling from a debt position reads as opportunism, and it retroactively poisons the complaint handling too, because now the customer suspects you were only being attentive to set up the ask. You lose the deal and the goodwill in one move.

What you do instead is close the complaint properly, commit to a specific follow-up with a date, and end the call. Then run the expansion conversation on a separate call once the fix has landed and the customer has acknowledged it. The gap costs you a week or two. It buys you the right to be heard.

The inverse also holds. If the account is quiet and healthy, a routine check-in is a much better vehicle for this than a formally scheduled "strategic review," because the formal meeting sets the audit expectation before you have opened your mouth. Turning an ordinary check-in into an expansion without setting off alarms is its own craft — the freight and 3PL playbook walks the version where the account is used to you calling about exceptions, and the staffing one handles the quarterly-review version where the customer arrives already braced.

When the numbers are genuinely bad

Sometimes there is no peak. Nobody is using it well. The spread is flat because everyone is at the floor.

Do not manufacture a bright spot. Customers can smell a curated data set, and the moment they catch you selecting for the flattering slice, everything else you say gets re-read as sales copy.

What you do instead is say the true thing, early, and give up the expansion. "I looked at the last quarter and I can't find a group that's really running with this. That's on us as much as anyone. Before we talk about anything else, I'd like to understand what happened after go-live." You will not sell anything on that call. You will find out whether there is an account left to sell to, which is more valuable and much rarer information.

And when you do eventually earn the expansion in that account, it will be because you were the vendor who said the uncomfortable thing first, instead of the one who arrived with a report.

Say it out loud before you say it to them

Everything above lives or dies on delivery. The framing sentence works when it sounds like a thought you just had and fails when it sounds like a line you memorized. The pause after the peak only works if you can physically tolerate the silence, and most reps cannot the first several times they try it.

So if I were picking up this approach tomorrow, I would not read it twice and dial. I would run the opening thirty seconds and the constraint question against a simulated buyer in DrillCall until the audit tone was gone from my voice, including the version where the customer opens with a complaint and I have to walk away from the pitch I prepared. That last rep is the one worth practicing, because it is the one you will be most tempted to skip when the number is close.