Upsell and Cross-Sell Scenarios: Roleplay for Existing Customer Growth

The Tuesday morning that exposes the gap

Picture a routine quarterly business review. An account manager joins a call with a customer two years into a mid-five-figure contract. The customer is nominally healthy: renewals have gone through, the champion answers emails, support tickets are unremarkable. Twenty-two minutes into the call, the customer mentions offhand that their operations team is standing up a new regional office next quarter and "we'll figure out the reporting thing later."

The rep says, "That's great, congrats." Then moves to the next agenda item.

That single sentence contained a budget cycle, a headcount expansion, a named unsolved problem, and an implicit deadline. It was the entire expansion opportunity, delivered voluntarily, and it evaporated because the rep had no trained reflex for what to do with it. This is the failure mode roleplay exists to fix — not the dramatic objection, not the competitive knife fight, but the quiet non-response to a signal that walked into the room on its own.

Run this exact moment as your first roleplay of any session. Give the customer-role player a script with one buried signal in it and instruct them not to repeat it or emphasize it. Give the rep no warning. Then debrief on whether the rep caught it, and if they caught it, what they did in the next thirty seconds. Most teams discover that catching the signal is not the hard part — roughly two-thirds of reps notice something was said. The breakdown is in the follow-through: the rep hears it, files it mentally as "something to bring up later," and later never arrives because there is no calendar invite, no CRM field, and no owner.

The scenario generalizes well past software. A commercial HVAC company hears "we're taking over the second floor." An accounting firm hears "we're setting up an entity in Texas." A staffing agency hears "our Q3 headcount plan got approved." Same structure, same missed handoff. What makes the roleplay valuable is that it isolates a behavior — signal-to-action conversion — that never gets isolated in real calls because real calls have twelve other things happening at once.

A useful variant: run the same scenario twice, back to back, with the same rep. The first pass is cold. The second pass, the rep knows the signal is coming. The gap between those two performances is exactly the gap that training closes, and showing a rep that gap on the same day is more persuasive than any benchmark you could quote them.

How the expansion motion actually works underneath the script

The reason expansion conversations feel different from new-business conversations is that the qualification burden has shifted, not disappeared. In a net-new deal you are establishing that a problem exists, that it costs money, that you are credible, and that budget can be found. In an expansion, the first three are largely settled — the customer already pays you, already trusts the category, already has an internal champion who defended the original purchase. What remains unsettled, and what reps consistently underweight, is that the economic buyer for the expansion is frequently a different person than the economic buyer for the original deal.

This is the mechanical heart of it. A customer buys a project management tool for the IT department. Two years later the expansion opportunity lives in Product, whose VP has never spoken to you, has no relationship equity with your company, and evaluates your proposal on cold merit alongside three competitors. Your champion in IT can introduce you but cannot authorize the spend. Reps who treat the expansion as "an easy conversation with a friend" walk into a room where the friend has no signature authority.

Qualification frameworks like MEDDPICC map cleanly onto this if you re-read each element through an expansion lens rather than reusing the new-business interpretation. Metrics become usage telemetry you already own — seat utilization, feature adoption, ticket volume, storage consumption — which is a genuine advantage over new business where you have to ask. Identify Pain becomes pain *expansion*: not "do you have a problem" but "the problem you already told us about has grown, here's the evidence in your own data." Competition becomes, in the overwhelming majority of expansion deals, the internal status quo — a customer saying "we're fine" is a competitor, and a formidable one, because doing nothing costs them nothing this quarter.

Decision Process is where expansion genuinely runs faster, and it's worth quantifying for your team so they calibrate. Amendments to existing contracts typically skip procurement entirely below a company-specific threshold; the legal review is lighter because the master agreement is already signed; security review is usually waived for an add-on module on an already-approved vendor. That structural advantage is real. It is also why reps get lazy — the mechanics being easier does not make the *persuasion* easier, and conflating the two is the single most common analytical error in expansion selling.

There's a downstream effect most training misses: the expansion motion changes who owns the account. If the customer success manager runs the discovery and the account executive runs the close, you have introduced a handoff into the middle of a trust relationship. Roleplay the handoff itself. Have the CSM introduce the AE on a live call and watch how many teams fumble it — the CSM over-explains, the AE opens with a pitch, and the customer visibly recalibrates from "my partner" to "being sold to." A clean handoff sounds like the CSM framing the AE as a specialist brought in to solve the specific problem the customer named, then going quiet.

What the numbers actually support, and where they don't

Be careful with expansion statistics. A great deal of what circulates as fact in sales training is vendor marketing that has been repeated until it acquired the texture of research. Two claims deserve particular skepticism, and you should not build your workshop on either without checking your own data.

The first is the acquisition-versus-retention cost ratio, usually stated as some multiple like five or seven to one. The underlying idea is sound — acquiring a customer costs materially more than growing one — but the specific multiple varies enormously by business model, sales motion, and how you allocate overhead. A product-led company with a self-serve funnel has a completely different ratio than an enterprise field-sales organization. Rather than quoting a number you cannot defend, compute your own: total sales and marketing spend attributable to new logos divided by new logos closed, versus total spend attributable to expansion divided by expansion bookings. Reps believe their own company's number in a way they never believe a slide.

The second is attach rate benchmarks. Published expansion or attach-rate figures depend entirely on how the reporting company defines an "attach" — whether it includes seat growth, whether it counts price increases, whether renewals-with-upgrade count as expansion. Two companies reporting the same headline number can be measuring different things. Use the definition to instrument your own CRM before you use anyone's benchmark to set a target.

Here is what you *can* measure reliably inside your own organization, and what makes for honest workshop targets:

Net revenue retention. The cleanest single expansion metric because it nets expansion against churn and contraction. Compute it monthly on a cohort basis, not blended, or growth in new cohorts will mask decay in old ones. If your NRR is under 100%, expansion roleplay is not your problem — retention is, and running upsell drills on an account base that's leaking is a way to feel productive while the floor gives out.

Time-to-expansion-mention. In a recorded QBR or discovery call, how many minutes elapse before the rep raises any growth topic? This is trivially measurable with a conversation intelligence tool or, honestly, with a stopwatch and a manager listening to five calls. It's the metric most directly moved by roleplay because it's a behavior, not an outcome.

Question-to-statement ratio in the expansion segment of the call. Count them. Reps who diagnose ask more than they assert; reps who pitch do the reverse. You don't need a benchmark for this — you need each rep's own ratio measured twice, before training and thirty days after, and the delta is the coaching conversation.

Objection recovery turns. When a customer pushes back, how many conversational exchanges pass before the rep is asking a question again rather than defending? Reps who never recover are the ones who quietly stop raising expansion topics at all, because the experience is unpleasant and nothing in their comp plan forces it.

Expansion opportunity creation rate per account per year. How many expansion opportunities does the average account generate, and what's the distribution? A long tail of accounts with zero expansion opportunities logged in eighteen months is either a real signal about product fit or a signal that nobody is looking. Roleplay fixes the second case and cannot fix the first.

Set targets from your own distribution. Take your top quartile of reps on expansion bookings, measure their behaviors, and make the median rep's target the top-quartile behavior — not an industry number that may not describe your business at all.

Choosing between the approaches, and what each one costs you

There is no single correct expansion motion, and picking one has real consequences for how you train. Four broad approaches show up in practice, each with a distinct roleplay implication.

Usage-triggered expansion. Telemetry fires an alert when an account crosses a threshold — seats near capacity, storage climbing, a feature adopted past some usage level — and a rep reaches out. This is the highest-signal approach and the easiest to scale, but it only works for products that generate meaningful usage data, and it biases toward selling *more of the same* rather than genuinely new capability. Roleplay implication: the rep must practice opening with data without sounding surveillance-creepy. There's a real line here, and reps cross it constantly. "I noticed you're at 94% of your license count" lands very differently from "our system flagged your account." Practice the phrasing.

Business-review-driven expansion. A scheduled strategic conversation, typically quarterly, where growth topics are on the agenda by design. Higher quality conversations, much lower volume, and expensive in rep time. Roleplay implication: the hard skill is agenda control — keeping a ninety-minute review from becoming a support escalation session, which is what customers will turn it into given the chance.

Renewal-adjacent expansion. Raising growth during or near the renewal conversation. Efficient in calendar terms and genuinely risky in practice, because renewals activate procurement, activate cost-scrutiny reflexes, and put the customer in a defensive posture where every proposal reads as a price increase. Many teams deliberately separate the two conversations by sixty to ninety days for exactly this reason. Roleplay implication: if you do run them together, practice the transition sentence, because that's where it breaks.

Service-led expansion. The CSM or support organization surfaces the opportunity through a genuine helping interaction — a training session, an unused-feature walkthrough, a workflow audit. Highest trust, slowest, and it requires a comp and incentive structure that doesn't punish CSMs for spending time on non-quota work. Roleplay implication: the challenge is the transition from helper to seller, and the honest answer is that it should often be a handoff rather than a transition.

Most organizations should run two of these, not four. Picking two lets you build genuine depth in the roleplay rather than skimming all of them. If you're unsure which two, pick the one your product data best supports and the one your team's existing calendar already contains — motions that require inventing new meetings tend not to survive a busy quarter.

A note on the persona dimension, which cuts across all four. A finance stakeholder, an engineering leader, and an operations manager evaluate the same expansion on genuinely different criteria — payback period and contract consolidation for one, implementation burden and integration risk for another, workflow and headcount effects for the third. Rotate personas during your workshop rather than letting each rep practice against the persona they find most comfortable. The comfortable persona is almost always the one they already sell to well.

Where these sessions go wrong

Roleplaying against a fictional account. Invented companies produce invented objections. Use real accounts from your own CRM, with real usage numbers and real open tickets, anonymized only as far as legal requires. The moment a rep can say "well, that wouldn't happen with my accounts," the exercise has lost its teeth.

Letting the customer-role player be too cooperative. Salespeople playing customers are pathologically helpful — they hand over the pain, agree to the next step, and everyone leaves feeling good. Give the customer role a written card with specific instructions: be distracted for the first two minutes, mention the signal once and don't repeat it, say "we're fine" the first time growth is raised. Constrained roles produce useful friction; free-form roles produce theater.

Debriefing on style instead of behavior. "You sounded confident" is not coaching. Score against a short checklist — did the rep ask a diagnostic question before proposing anything, did they reference a specific number, did they identify who signs, did they end with a dated next step — and let the checklist carry the feedback. Four to six binary items, no more. Longer rubrics get abandoned by week three.

Skipping the objection block. Teams love the discovery portion and skip objections because objections are uncomfortable to practice. That's precisely why they need practice. Pull your three most common expansion objections from actual CRM loss reasons — not from a generic list — and drill each one for a few minutes with a rotating cast. The objection you skip in the workshop is the one that kills the deal in October.

Assuming the workshop is the intervention. A single session decays fast. What sustains behavior change is the follow-up loop: each rep brings one real recorded call thirty days later and self-critiques it against the same checklist the workshop used. This is the part most organizations drop, and dropping it makes the workshop a morale event rather than a training one.

Ignoring the CRM plumbing. If there's no clean way to log an expansion opportunity distinctly from a new-business opportunity, your pipeline reporting will silently blend them and you will never be able to tell whether any of this worked. Create the opportunity type, link it to the parent account, add fields for the triggering signal and the identified economic buyer, and build one report. Do this before the workshop, not after — reps who leave a session with nowhere to put their commitments don't keep them.

Training expansion into an account base that's churning. Worth restating as a pitfall in its own right. Expansion motion on unhealthy accounts reads as tone-deaf to customers and demoralizes reps who are being asked to sell more to people who are unhappy. Audit account health first. If a rep's book is in trouble, their assignment coming out of the workshop should be retention conversations, not growth ones, and saying so out loud protects the credibility of the whole program.

Related questions

How long should an expansion roleplay session run?

Ninety minutes is a practical ceiling for a full session with two scenarios and a debrief. Beyond that, attention collapses and the last exercise gets rushed. Shorter fifteen-minute single-scenario drills at the start of a weekly standup are often more effective because they recur.

Should customer success managers run expansion roleplay too?

Yes, but with different scenarios. CSM roleplay should focus on surfacing and qualifying the opportunity, then handing off cleanly — not on negotiating or closing. Practicing the handoff sentence itself is the highest-value fifteen minutes a CSM team can spend.

What if we don't have call recording software?

You don't need it to start. A manager sitting in on five live calls per rep per month, with a written checklist, produces most of the coaching value. Recording tools add scale and searchability, not fundamentally new insight.

How do we roleplay expansion for a product with no usage telemetry?

Substitute other observable signals: support ticket themes, invoice line items, org chart changes, publicly announced customer initiatives. Services businesses run expansion on these exclusively and do it well. The discipline is the same — anchor the conversation to something specific you observed.

Does expansion roleplay work for transactional, low-ACV accounts?

Partially. For very high-volume, low-value accounts the leverage is usually in product-led prompts and lifecycle messaging rather than rep conversation. Reserve human roleplay investment for the account tier where a conversation actually changes the outcome.

FAQ

How do we handle the "we're fine right now" response?

Treat it as a real position rather than a brush-off, because usually it is one. The productive move is to accept it and narrow: acknowledge that things are working, then ask a specific question about the one area where you have evidence of friction. If the customer confirms the friction is real but not urgent, get a date rather than a decision — a specific month to revisit, tied to something already on their calendar.

Should the same person handle renewal and expansion?

It depends on your comp structure more than your org chart. If the renewal owner is measured on retention alone, they will avoid raising growth topics that risk the renewal. If one person owns both, make sure the incentive weights don't make expansion a rounding error against the renewal number. Many teams split the conversations by date rather than by person, which preserves the relationship while separating the psychological frames.

How much of the workshop should be practice versus instruction?

Heavily weighted toward practice. A reasonable split is fifteen minutes of framing, sixty minutes of active roleplay and debrief, and fifteen minutes of individual commitments. Framework instruction beyond that is better delivered as pre-reading. Reps do not need another explanation of a qualification framework; they need repetitions under mild pressure.

What's the right group size?

Three to four people per breakout — one rep, one customer, one or two observers — with rotation so everyone plays each role. Larger groups mean less air time per person and more spectating. Playing the customer is unexpectedly instructive; reps who have sat in the buyer's chair make better diagnostic openings afterward.

How do we keep this from feeling like theater?

Real accounts, constrained customer roles, and a checklist-based debrief. Also: the manager should go first and do it badly in front of everyone. The single biggest barrier to useful roleplay is that people are afraid of looking incompetent in front of peers, and the only reliable fix is for the most senior person in the room to absorb that risk first.

What should reps leave the session with?

One named account, one specific signal they'll reference, one identified economic buyer, and a date. Written down, logged in the CRM, and reviewed at the next standup with the question "did you make the call, and what happened?" Commitments without a review date have roughly the durability of a New Year's resolution.

Sources

• Gong Labs — sales conversation research
• Harvard Business Review — sales and account management research
• Salesforce — CPQ and contract amendment documentation
• SaaS Capital — net revenue retention benchmark research
• OpenView Partners — SaaS benchmarks and expansion metrics
• Bain & Company — customer loyalty and retention research
• Winning by Design — revenue architecture resources
• Gartner — sales and customer service research

Related on PULSE

• [60-Min Sales Training: Cross-Sell to Existing Customers](/knowledge/st0475)
• [Top 10 Upsell and Cross-Sell Templates for Team Training](/knowledge/st0737)
• [The Value-Add Upsell: Template for Identifying Cross-Sell Opportunities in Accounts](/knowledge/st0699)
• [The Cross-Sell and Upsell Reboot — 60-Min Training](/knowledge/st166)
• [Objection Handling Roleplay Session Blueprint](/knowledge/st0752)
• [Deal Desk Simulation: Cross-Functional Approval Process Roleplay](/knowledge/st0710)