CSM Expansion Playbook: Uncovering Upsell Without Becoming Sales

Two CSMs work at the same company. They have similar books of business, similar tenure, similar product knowledge.

The first one never raises pricing or seats. She believes commercial conversations belong to sales, full stop. Her customers love her. They also leave at renewal because their contracts no longer match what they're using, and nobody on the vendor side connected the dots in time. Her NRR is 94%.

The second one pitches an upgrade in every QBR. He has a slide deck and a price list ready before the customer has even said what's working. His customers are polite about it the first two times. By the third QBR, the buyer has quietly added a sales filter to inbound emails and started routing strategic questions to the support inbox instead. His NRR is 102% but his churn is climbing.

Both are missing the actual job. The CSM's job in expansion isn't to avoid the conversation, and it isn't to pitch the upgrade. It's to help the customer find the next outcome worth paying for, and then bring in the right person to close the commercial mechanics.

That middle path is the entire skill, and most CSM training programs don't teach it.

Why Expansion Is the CSM's Job, Not Sales'

Net Revenue Retention above 120% (the bar most boards expect from healthy SaaS) is mostly expansion, not new logos. Gross retention plus 20+ points of expansion is what gets you there. New logos don't.

And the expansion that compounds quarter over quarter doesn't come from outbound prospecting into the install base. It comes from CSMs surfacing real outcomes the customer is ready to pay for, because the CSM is the only person on the vendor side who actually knows what the customer is doing with the product day to day.

The trap I fell into early was thinking "expansion" and "upsell" were the same word. They aren't. Upsell is a sales motion: prospect, pitch, close. Expansion is a customer-success motion: surface a need the customer has, frame the outcome, then hand the commercial mechanics to someone whose job it is to negotiate price and paper. When you collapse the two, you either avoid the topic entirely (and lose NRR at renewal) or you pitch like an AE and damage the relationship that made you valuable in the first place.

Pressure-led upsell motions close faster on the first deal and slower on every subsequent one, because the customer learns what your meetings are actually about. Trust-led expansion compounds because the customer is the one driving the conversation forward. The numbers track this. Most CS leadership benchmarks I've seen put CSM-sourced expansion deals at 2-3x the close rate of cold AE-sourced expansion into the same accounts.

So the question isn't whether CSMs should drive expansion. It's how. And the answer is built around three things: a scoring rubric so you can spot the signals before you act, a conversation script that surfaces the need without pitching, and a handoff that hands commercial mechanics to sales without breaking trust.

The Expansion-Signal Scoring Rubric

You need a system. Without one, you'll either chase every uptick in usage as if it's a deal, or you'll miss the customer who has hinted three times that they need more.

Here's the rubric we use. Five signals, point-weighted, with a clear threshold for action.

Threshold for action: 5 points.

A 5-point account gets an expansion-conversation slot in the next QBR. A 7-point account gets one within two weeks, regardless of QBR cadence. A 3-point account gets a note in the CRM and a re-score in 30 days.

The rubric does two things. It stops you from acting on every twitch in usage data, and it forces you to write down the evidence before you bring it to anyone. When you eventually loop in sales, you're not saying "I have a hunch." You're saying "this account is at 7 points, here's the breakdown, here's the customer language behind each signal."

That's the difference between a CSM who sources pipeline and a CSM whose pipeline gets ignored.

A note on what's NOT on the rubric: NPS score, support ticket volume, executive engagement frequency. Those are health signals, not expansion signals. Healthy accounts don't automatically expand. They renew. Expansion needs a specific gap between what the customer bought and what they now need. Health tells you whether the conversation is safe to have, not whether there's a conversation to have at all.

The Customer-Led Expansion Conversation

Once an account crosses the threshold, you book the conversation. Not as an "upsell call." As a working session about what the customer is trying to accomplish next.

The shape of the conversation is the same every time: Open → Mirror → Constraint → Unlock → Next Step. Each step has verbatim language I've used in real calls. Adjust the words to your voice, but don't skip a step.

Open: Ask about their goals first

"Before we get into product specifics, I want to make sure I understand what you're trying to accomplish in the next two quarters. Walk me through what 'good' looks like for your team by end of Q3."

You're not pitching anything. You're letting them set the frame. Most customers will spend 5-10 minutes here, and what they say is the most valuable thing in the meeting. Take notes. Don't interrupt to draw connections to the product yet. You will be tempted. Don't.

Mirror: Reflect back what's working

"What I'm hearing is that the [team/workflow/use case] we set up in Q1 is now the way you actually run [process]. That tracks with what I'm seeing in usage. Your team is at 84% of contracted seats and the [feature] adoption is up 60% from launch. Did I get that right?"

This step matters because it confirms you've been paying attention to their reality, not just running a script. It also surfaces specific evidence from your scoring rubric without naming the rubric. The customer should hear "you noticed" rather than "you have a system."

Constraint: Name the gap they're hitting

"The thing I'm watching, and I want your read on this. When [adjacent team] starts using the workflow you've built, you're going to hit your seat ceiling around mid-Q3. You'll either have to triage who keeps access, or you'll have to expand the contract. Is that something you've started thinking about?"

This is the pivot. You're not pitching expansion. You're naming a constraint they will face, with specific evidence, and asking them to confirm or correct your read. If you got the signals right, they'll nod. If you got it wrong, they'll tell you, and you've just learned something more valuable than a deal.

The rule for this step: never name a constraint you don't have evidence for. "I noticed you're at 84% utilization" is fine. "I think you might need more seats" is not. That's pitching.

Unlock: Ask what solving it would mean

"If we figured out how to clear that constraint cleanly, what would it unlock for the team? What would you be able to do that you can't do today?"

This is the question most CSMs skip, and it's the one that decides whether the deal closes well or limps. The customer's answer becomes the business case the AE will use later. It also tells you whether the expansion is actually worth doing. Sometimes the unlock is small enough that the customer doesn't need to expand, just optimize. That's a healthier outcome than a deal that doesn't deliver.

If the unlock is real and material, you've earned the next step.

Next step: Bring in the right person

"There are two paths from here. One is we figure out a workaround that fits your current contract, and I'll walk you through that. The other is we look at what an expansion looks like, in which case I'll bring in [AE name] for the commercial side and stay in the meeting myself so we keep the context. Which feels right to you?"

You're handing them control. You're also signaling, clearly, that you are not the commercial decision-maker, but you are still accountable to the relationship. That framing is what separates a trusted CSM from a CSM who just got asked to leave the room.

The Warm Sales Handoff Template

The handoff is where most CS-sourced deals die. The CSM finds the signal, runs the conversation, gets the customer warm, and then drops the AE in cold like a stranger. The customer recoils. The AE, who hasn't been briefed, asks discovery questions the CSM has already answered. The deal stalls.

The fix is a written handoff document the AE reads before the call, plus a clear role split during it.

Expansion handoff template:

ACCOUNT: [Customer name]
SCORE: [X points], [date scored]
CSM: [Your name]
AE: [AE name]
HANDOFF DATE: [Date]

1. ACCOUNT CONTEXT (3 sentences max)
What does the customer use us for today, what team is the buyer in,
who is the executive sponsor.

2. SIGNAL EVIDENCE
- [Signal 1 + specific data point]
- [Signal 2 + specific data point]
- [Signal 3 + specific data point]
Pull from the scoring rubric. No vibes, only evidence.

3. CUSTOMER LANGUAGE
What did the customer actually say in the conversation? Quote them.
The AE will use these phrases back at them. If you paraphrase, the AE
will too, and the customer will hear themselves being misquoted.
- "[Verbatim quote about their goal]"
- "[Verbatim quote about the constraint]"
- "[Verbatim quote about what solving it would unlock]"

4. COMMERCIAL RANGE
What is the likely shape of the deal? Seats, modules, contract length.
Don't guess at price; that's the AE's job. But give them the units.
- "Probably +15-25 seats, possibly the [module] add-on, willing to
  discuss multi-year if it pencils."

5. CSM'S CONTINUED ROLE
What is the CSM doing in the deal cycle? In the meetings? Post-close?
- "I'll be in every commercial meeting as the trusted voice. AE owns
  pricing, paper, procurement. I own the relationship and the
  outcome story. Post-close, I run the expansion onboarding."

6. THINGS THE AE SHOULDN'T DO
Anything specific to this account that would damage trust if the AE
defaulted to standard playbook? List it.
- "Don't lead with the year-end discount. The buyer hates feeling
  rushed and will read it as pressure."
- "The exec sponsor is new. Don't reference the original deal; she
  wasn't there for it."

The handoff happens at least 48 hours before the AE's first call with the customer. Anything less and the AE is reading it in the parking lot, which means they're not reading it.

During the call, the role split is clean. The AE drives the commercial conversation: pricing, terms, contract length, procurement timeline. The CSM stays in the meeting as the continuity voice and as the person who can confirm the outcome story when the buyer pressure-tests it. ("Camellia, we'd be moving from 50 seats to 75. Does that actually solve the constraint we talked about?" "Yes, here's why, and here's what we'll measure post-rollout.")

The CSM does not negotiate price. The AE does not own the relationship. Both stay accountable to the customer's outcome.

Common Pitfalls

A short list of the ways this goes wrong, in roughly the order I've seen them most often.

Acting like a quota-bag AE. Talking price before value. Bringing slides into a working session. Asking "are you the decision-maker?" to a champion who has been a champion for two years. Customers can smell a pitch shift the moment it happens, and they downgrade you in their head from "trusted advisor" to "vendor."

Missing the buying signal entirely. The customer hints three times that the marketing team wants in. The CSM says "oh interesting" and changes the subject. Six months later marketing buys a competing tool. The signal was there, the rubric would have caught it, but the CSM was operating on intuition and the intuition said "don't be salesy." The cure isn't to be salesy. The cure is the rubric.

No warm handoff. The CSM books the AE into the next call without briefing them. The AE shows up cold. The customer is now in a meeting with a stranger who is asking questions the CSM has already answered. The deal slows by 30 days minimum. Half the time it dies.

Pitching expansion before the current scope is delivering. If the customer hasn't gotten value from what they already bought, no signal score in the world makes expansion appropriate. Score the account, see the points, then check: is the customer actually getting outcomes today? If not, the next conversation is about delivery, not expansion. The rubric tells you when there's a deal. Health tells you whether to act on it.

Confusing renewal with expansion. Renewal is keeping what they have. Expansion is selling them something more. CSMs who blur the two end up doing renewal-flavored expansion conversations ("we should also talk about adding seats") that customers experience as a hostage negotiation. Renew first, expand second, and run the conversations on different days when possible.

Measuring Whether This Is Working

You need outcome metrics for the expansion practice itself, not just the deals that close. Four to track:

• Expansion conversations per quarter, per CSM. This is an activity metric, but it's the one that predicts everything else. CSMs who run zero expansion conversations have zero expansion attribution. Aim for 1-2 per CSM per quarter for accounts above the action threshold.
• NRR contribution attributable to CSM-sourced opportunities. Tag deals at handoff. At end of quarter, look at what closed from CSM-sourced pipeline vs. AE-sourced. CS-sourced should be 40%+ of expansion bookings in a healthy team.
• Close rate of CS-sourced vs. AE-sourced deals. CS-sourced should close at 2-3x the rate of cold AE-sourced expansion into the same accounts. If they don't, your handoff is broken or your scoring is too loose.
• Customer satisfaction post-expansion conversation. Send a 1-question survey 14 days after any expansion conversation, deal or no deal: "After our last working session, how would you rate the conversation?" Trust shouldn't drop. If it does, even on deals that closed, the CSM ran a pitch, not a conversation.

The fourth metric is the one most CS orgs don't track and the one I'd argue matters most. Closed deals are easy to count. Trust eroded by a clumsy expansion attempt is invisible until renewal, and by then it's too late.

How Rework Supports This Workflow

The expansion-signal scoring rubric only works if the data is in front of you when you're scoring, and if the handoff document survives the trip from CSM to AE. Rework Work Ops keeps both surfaces in one place: score accounts using the rubric on each customer's profile, attach the evidence (usage screenshots, customer quotes, support ticket links) directly to the score, and turn the score into a handoff document that the AE actually opens. Rework CRM syncs the handoff and customer language into the deal record, so the AE walks into the call with the same context the CSM has, not a different version of it. CRM starts at $12/user/month, Work Ops at $6/user/month. Most CS teams run both, since the relationship side and the deal side need to share state.

The Honest Summary

Trust enables expansion. Pressure kills it. The CSM's job is to surface the opportunity, frame the outcome with the customer's own language, and hand commercial mechanics to sales while staying in the meeting and accountable to the relationship.

The handoff is the make-or-break moment. Not the pitch, not the deck, not the discount. The handoff. If you brief the AE properly, the AE doesn't show up cold and the customer doesn't feel sold to. If you don't, the deal dies and the relationship gets a small scar that will show up at renewal.

If you run expansion this way for a year, you'll find that your best customers start asking you about expansion before you ask them. That's the goal. The CSM who never raises commercial topics loses NRR at renewal. The CSM who pitches in every QBR loses access to the room. The CSM who scores systematically, runs customer-led conversations, and hands off warmly ends up with a book of business that grows itself.

That's the playbook. The rest is reps.

Learn More

• Customer Onboarding: The First 30 Days That Decide NRR
• QBRs Customers Actually Look Forward To
• CSM Metrics That Matter: NRR, GRR, and Health Scores
• The CSM to Senior CSM Path
• Customer Success Manager Job Description