Sales Voicemail Practice: A Scored Drill Guide

Why does sales voicemail practice belong before live dials?

Because an improvised voicemail uses a live prospect as the rep's first draft. Rehearsal moves that draft into a scored session, where a miss costs a re-run instead of a contact. We hold that reps improve fastest by practicing high-stakes conversations before facing buyers, and a voicemail is a short one that is easy to score.

It is also a straightforward outbound conversation to verify. The rep speaks alone, the message is short, and the transcript holds the words the listener would hear. A manager can check every row of the rubric from text, with no buyer reaction to interpret.

The opposing view is that reps learn voicemail by leaving a lot of them. Live volume does give repetitions. It does not give a transcript scored against a named row, and without that, nobody has named which row the rep keeps missing. We recommend rehearsal alongside live dialing, not instead of it.

A voicemail transcript holds every word the listener will hear, which makes each rubric row checkable without a recording.

What should an outbound voicemail scoring rubric check?

Four rows: a stated reason for calling, one specific ask, a callback number said clearly, and a time-box. Each row is an observable behavior with pass evidence a manager can find in the transcript. None of them requires hearing the rep's tone.

If your methodology or call stages name the opening differently, rename the rows to match your playbook. This publication holds that roleplay scoring changes behavior only when it is graded against your own methodology and call stages, and that a generic rubric produces feedback reps rationally discount. Treat that as our position, and test it against your own team's transcripts.

We set the time-box at a 30-second ceiling. That is our recommendation, not a finding. Pick your own cap before the session starts and do not change it mid-drill.

What does a passing take sound like, and a failing one?

A passing take names the reason in the first breath, makes one request, and says the number twice. A failing take misses one or more of those rows. The worked scenario: a discovery call ended Tuesday, the follow-up call went to voicemail, and the rep has one ask.

Passing take: Hi Dana, it's Sam Ortiz at Northwind. I'm calling because on Tuesday you said the manual handoff between sales and onboarding is slowing new accounts, and I have one change to test with you. Can you reply to this message with a yes or no on a 15-minute call Thursday morning? My number is [number, spoken in groups, with pauses]. Again, [number]. Thanks, Dana.

Failing take: Hey Dana, Sam from Northwind, just following up and checking in on my last email. We help companies like yours with workflow, reporting, integrations, a bunch of things, so give me a call back or check out the website, or I'll try you again next week, my number's [number, rushed]. Talk soon.

Score the failing take row by row. Reason: fail, because checking in names no trigger. Ask: fail, because there are two and neither has a time. Number: fail, because it appears once and fast. Time-box: fail, because the feature list pushed it past the cap. Four rows failed, and the debrief picks one.

A rubric row counts only if a manager can verify it from the transcript without hearing the rep's tone.

How do you run the voicemail drill in 30 minutes?

Run it as 5 minutes of setup, 15 minutes of drill, and 10 minutes of debrief. That is the run cost for the manager. Scenario prep happens beforehand: about 10 minutes to pick the scenario from a live deal and choose the rubric row you will watch hardest. This voicemail drill for sales reps works on any methodology because it scores behaviors, not model vocabulary.

Setup (5 minutes): hand the rep a scenario card with the buyer's name and role, what happened on the last contact, and the one ask. State the cap and the pass bar aloud. Drill (15 minutes): the rep records up to three takes. Each take is recorded, timed, and transcribed. On the final take, no notes are in view. In XL Roleplay, sessions are recorded, timed, and transcribed, so the timestamps and the number check come from the session record. A stopwatch and a transcript do the same job.

Pass bar: all four rows pass on the same take. A take that passes three rows is a fail. Failure is not scored as partial credit and not marked complete.

Debrief (10 minutes): follow the one-behavior format in our sales roleplay debrief guide. The rep reads the transcript and names the failing row first. The manager then confirms or corrects, quoting the transcript line. If two rows failed, pick one for the debrief. The re-run is still scored on all four. Book the re-run date before the rep leaves.

A voicemail drill passes only when all four rows hold on the same take, and a re-run is booked before the debrief ends.

Where does a scripted first line sound stiff?

Expect stiffness in the first sentence, where a scripted reason for calling gets recited instead of said. The trade-off is real: the rubric rewards structure, and structure delivered flat still passes. Our judgment is that a cold call voicemail script should not be fixed word for word.

Our recommendation is to agree on the frame, not the sentence. Script the opening shape (I'm calling because plus the trigger), the single ask, and the number sequence. Leave the trigger itself in the rep's own words, since it comes from the real deal. The failing take above is stiff in a different way, with filler where the reason should be. Our piece on script practice without sounding scripted covers how to rehearse the frame until it loosens.

Three other trade-offs. A transcript cannot score warmth, so a take can pass all four rows and still sound cold; that is a coaching conversation, not a rubric row. The 30-second cap may be too tight for some offers, so set your own and hold it. Saying the number twice costs seconds against the cap. We think the cost is worth paying, but you may decide otherwise.

Finally, the pass bar certifies behavior. It says nothing about whether any buyer calls back, and we have no callback data to offer.

What drill can you run this week?

Run one 30-minute session on one open follow-up where the last call went unanswered. Keep it separate from your pipeline 1:1, as an addition to it. For multi-touch follow-up after a ghosted prospect, see our ghosted prospect drill. For a live opening, see the call opening drills.

Before the session (about 10 minutes): pick the deal, write the scenario card, and choose the row you expect to fail. A callback message roleplay on one real open deal is enough for the first week.

In the session: 5 minutes setup, 15 minutes of up to three recorded takes, 10 minutes debrief. Cap: 30 seconds. Pass bar: reason, ask, number, and time-box all pass on one take with no notes in view.

A failing attempt sounds like just checking in, an ask that offers two ways to respond, a number said once at speed, or a message that runs into a feature list. On any fail, the rep names the failing row from the transcript, you confirm with the line, and you book the re-run inside the same week. The rep leaves a live voicemail on that deal only after a passing take.

Next week, read the transcripts per rep and per scenario. Do not blend them into one team score.