GuidesPublished 9 min read

Referral Request Roleplay: Drill the Intro Ask

two forms meeting in a spotlight pool
Listen to this article · 11:35 · AI-generated narration
0:00 / 11:35
Chapters

TL;DR

A referral request roleplay drills one thing: leaving the conversation with a named person, a reason for the introduction, an agreed mechanism, and a date. Trigger the ask after the customer states a result in their own words, open it with a short mutual agreement, and rehearse first responses to three common stalls. Score the attempt on four rubric rows and re-run any attempt that ends without a name on the transcript.

  • Trigger on a stated result, not on contract signature or renewal paperwork.
  • Bring named candidates to the ask; broad questions produce broad non-answers.
  • Drill one agreed first response to each of the three common stalls.
  • Pass bar: named person, stated reason, agreed mechanism, dated follow-up - all four visible in the transcript.
  • Failed attempts get re-run the same week, not logged as completed.

When should the referral ask happen in the account lifecycle?

Ask after the customer has said out loud that something worked. Not at signature, when value is still a promise. Not during an escalation, when goodwill is overdrawn. The trigger you want is a verbatim moment you can point to: a quarterly business review where the sponsor described a result, a support save that the customer thanked your team for, a renewal where the champion volunteered a reason to stay.

Ask for an introduction after the customer has stated a result in their own words, not after they have signed.

That timing rule gives you the second half of the ask for free. You are not asking a favor in a vacuum; you are asking someone who just articulated an outcome whether another person with the same problem should hear about it. Write the trigger into the account plan the same way you write a renewal date. If your team runs structured reviews, the referral ask belongs on the agenda of the review that follows a confirmed win, and the person who owns the account owns the ask.

There is a fair counter-argument: referrals sometimes appear spontaneously, and a rigid trigger can make reps skip organic openings. Take both. Keep the scheduled trigger so the ask happens at least once per healthy account per year, and coach reps to take any unprompted opening the customer hands them. What you are eliminating is the pattern where the ask never happens because no moment ever felt right.

Open the ask with a mutual agreement, then name a person

Two mechanics separate an ask that lands from one that evaporates. The first is a Sandler up-front contract on the ask itself: purpose, time, both agendas, and permission for the customer to decline. It sounds like this on the transcript: "Before we wrap, can I take three minutes on something that isn't about your account? If the answer ends up being no, that's a fine answer and nothing changes on our side."

The second is specificity. The hard part of asking for referrals in sales is not courage; it is arriving with names. Reps who say "do you know anyone who'd benefit" are outsourcing research to the customer, and the customer's honest answer is that nobody comes to mind. Do the work first. Pull two or three named people from the champion's former employer, their peer group, or the sister business unit that sat in on the pilot, and ask about one: "I saw you worked with Dana Reyes at Northwind. Her ops team has the routing problem you had last spring. Would you introduce me?"

Bring the name yourself rather than asking the customer to produce one.

Notice what the rep has already banked with that one sentence: a name, and a reason for the introduction stated in the buyer's problem language rather than product language. Those are two of the four things the rubric scores. Both are prepared before the call, which is why this drill rewards preparation more than charm.

Verbatim first responses to the three stalls

We recommend drilling three common stalls. Agree on one first response to each, drill it until it is automatic, and let reps improvise the branches from there. Our position is that rehearsal is what makes improvisation safe - strong reps then improvise the branches from understanding instead of being caught flat.

Stall 1 - "Let me check with them." First response: "That's fair. What would you want to check with her?" Then close the mechanism and the date in the same breath: "If she's open to it, would you rather forward a note I draft, or should I reach out and mention you? Either way, can I check back Thursday?"

Stall 2 - "I'd rather not use my name." First response is a Sandler negative reverse - soften toward the resistance so the customer argues the truth: "Makes sense. Maybe an introduction isn't the right move here. Is it the timing, or is it that you're not comfortable vouching for the work yet?" One of those two answers is coaching you need. The other usually produces a smaller version of the introduction the customer will make.

Stall 3 - "Nobody comes to mind." First response: "That's usually because the question is too broad. Let me narrow it - who ran the evaluation with you at your last company?" Narrow by role, by company, by event. Never re-ask the broad version louder.

Every stall response ends by proposing a mechanism and a date, never by thanking the customer and moving on.

The four-row rubric for scoring the ask

Score the champion introduction request on four rows only. Four rows fit in one debrief and force the rep to pass all of them rather than trade one for another. Treat the four as referral ask exit criteria in the ordinary sense: buyer-verified evidence that must exist on the transcript before the attempt counts, not the rep's impression that it went well.

A referral is not earned until a specific person, a reason, a mechanism, and a date all exist in one transcript.

Grade the rows against your own call stages and account-management standards rather than a generic template. If your organization routes introductions through customer success, the mechanism row should say so, and a rep who agrees to a mechanism your company does not actually run should fail that row. Generic rubric language is the fastest way to teach reps that the score is theater.

Rubric rowPass evidence in transcriptFailing attempt sounds like
Named personA specific individual is named by the rep or the customer, with a company or role attached.Anyone in your network who might be a fit.
Stated reason for the introRep states, in the buyer's problem language, why this person should hear from them.I think they'd find our platform interesting.
Agreed intro mechanismCustomer agrees to forward a drafted note, send a three-way email, or permit a named mention.Send me something and I'll pass it along.
Dated follow-upA specific weekday or date is agreed out loud, owned by the rep.No rush, whenever you get to it.

Why do reps accept "send me something and I'll pass it along"?

Because it feels like a yes and it ends the discomfort. The customer sounds cooperative, the rep gets to stop asking, and both parties leave the call believing something was agreed. Nothing was. There is no named recipient, no visible send, no date, and no way for the manager to inspect what happened.

Be fair to the customer here. Sometimes routing a note through the champion genuinely is the right mechanism - their relationship, their words, their timing. The problem is not the forward. The problem is the missing name and the missing date. So do not refuse the offer; attach the two things it lacks: "Happy to draft it - I'll have it to you tomorrow morning. Who are you planning to send it to, and can I check in Friday to see whether it landed?"

A forwarded note with no recipient and no date is a polite decline the rep recorded as progress.

This is also where managers overrate confident reps. A rep who narrates the call as "great conversation, they're going to introduce me" is describing a feeling. The transcript either contains a name and a date or it does not. Score the transcript. If the same rep collects three of these soft yeses in a month, the pattern is a skill gap in closing the mechanism, and it belongs in a coaching block rather than a pipeline review. We keep those two meetings separate for exactly this reason - see Sales Coaching vs Deal Review.

The 15-minute weekly drill, its pass bar, and its re-run

Run it as a standing block. We recommend a manager run cost of 30 minutes - 5 minutes of setup, 15 minutes of drilling, 10 minutes of debrief - plus roughly 10 minutes beforehand to pick the account and choose the rubric rows. Be honest about that prep; the drill is only as good as the scenario you feed it.

Setup: choose a real account where the customer recently stated a result. The rep arrives with two named introduction candidates and one sentence of reason for each. No names, no drill - send them to prepare and re-book.

Drill: three attempts of about four minutes. Attempt one, the buyer persona cooperates. Attempt two, the persona uses one of the three stalls. Attempt three, the persona uses a different stall and adds mild time pressure. In practice you can run this against a peer playing the champion or against an AI persona in XL Roleplay, where the session is transcribed and each rubric row is flagged back to the exact moment it was won or lost.

Pass bar: in two of three attempts, all four rubric rows appear in the transcript. A failing attempt sounds like this: "If you think of anyone, keep us in mind - no pressure." Silence. "Great, I'll leave it with you." No name, no mechanism, no date.

Debrief one behavior, let the rep self-diagnose first, and schedule the re-run before anyone leaves the room. Failed attempts are re-run within the same week. The debrief structure we use is in Sales Roleplay Debrief: One Behavior, Re-Run, and the upstream work of building champions worth asking is in Sales Champion Development Drills.

Frequently asked questions

How many referral candidates should a rep bring to the ask?

We recommend two or three named people, researched before the call. One name makes the ask brittle if the customer has no relationship with that person; a long list turns the ask back into a memory test for the customer.

Should customer success or the account executive own the ask?

Whoever owns the relationship at the moment of the trigger should ask, and that owner should be named in your rubric's mechanism row. What breaks teams is ambiguity - two owners means no one asks and no one is accountable for the follow-up date.

Does drilling an agreed first response make reps sound scripted?

Only if they stop at the script. The first response removes the flinch; the branches after it come from listening. Reps who improvise the opening line under pressure usually default to the broad question that produces nothing.

What if the customer's company policy forbids references or introductions?

Score the attempt, not the outcome. A rep who named a person, stated a reason, and surfaced the policy constraint passed the drill; the account simply has a real blocker. Record the constraint so nobody re-asks the same question next quarter.

How often should this drill run?

Weekly while the behavior is new, then monthly as maintenance, scheduled as a protected coaching block separate from pipeline review. Referral opportunities are scarce, so rehearsing before the live ask costs far less than spending a real champion relationship on practice.

All insights