Practice cold calling without real prospects: a 4-week plan
A four-week plan to practise cold calling without burning real prospects: what to drill each week, how many calls a day, and when a rep is ready.
You can practise cold calling without a single real prospect on the line, and new reps should. Three formats do the job: a colleague plays the prospect, the team listens to recordings of real calls, or the rep dials an AI prospect that pushes back and scores the call. The mistake is treating them as interchangeable or doing one of them twice a month. A new rep needs around a hundred practice calls in four weeks, most of them on the first thirty seconds, before anyone in the pipeline hears them. The plan below is the one we have seen hold on real floors: what to drill each week, how many calls, which level of difficulty and what "ready" looks like in numbers.
Why practising on the pipeline costs more than it looks
The default way to learn cold calling is to make cold calls. The rep gets a login to the dialler in week two and starts working the list. It feels efficient because the activity numbers appear on a dashboard from day ten.
The cost is in the list. Every account a new rep fumbles in month one is an account somebody could have booked in month four, and the rep learns nothing from the fumble because nobody heard it. The second cost is in habits. A rep who learns to dial fast before they can hold a conversation keeps dialling fast for the rest of their tenure.
So the first four weeks separate practice from production. Practice is private, repetitive and scored. Production starts when the scores say so.
The three ways to practise, and what each one is for
A colleague plays the prospect. The only format where a human can stop the call halfway through an objection and say "that word, 'solution', is why I switched off, try again". Use it for judgement: when to push, when to let a prospect go. Its weakness is arithmetic. One rep, one manager, twenty minutes, perhaps three attempts.
The team reviews recorded calls. The only format where the prospect is real. A new rep hears a logistics VP say "we've already got a provider for that, and honestly it's fine" in the VP's own words. Review builds knowledge of what your market says and in what order. It does not build the habit of replying under pressure, because the call is over and the rep cannot try a different line.
The rep dials an AI prospect. The only format that gives a rep the same thirty seconds fifty times in a week with nobody watching. The better tools build the prospect from the rep's real target accounts, set resistance by level and score every call on the same rubric. The weakness is that it cannot explain. It produces the evidence for a coaching conversation and does not have the conversation.
We compared the three in detail in our piece on cold call simulators. The short version is that review tells you what the market says, the simulator gives you the repetitions, and roleplay is where the manager spends their twenty minutes on the rep the scores point to.
The four-week plan
The plan assumes a rep can make fifteen to twenty practice calls a day in about ninety minutes, which is what we see on teams using a simulator, and one twenty-minute roleplay with the manager per week. Levels refer to the three prospect types most simulators and sensible roleplay partners use: a warm prospect with time who raises one soft objection, a busy inbound lead who is interested but sceptical, and a cold senior buyer who was interrupted mid-task and tries to end the call twice.
| Week | Drill | Calls per day | Level | Ready to move on when |
|---|---|---|---|---|
| 1 | Opener and reason for the call only. Hang up after the prospect responds. | 20 | Warm | Rep says name, company and a specific reason for calling in under fifteen seconds without reading, twenty times in a row |
| 2 | Opener, reason, one objection. Pick the team's top three objections and run each one. | 20 | Warm, then inbound from Wednesday | Opener averages 7 or better over three calls at the inbound level. Rep acknowledges the objection before answering, every time |
| 3 | Full call, with two discovery questions and a close. | 15 | Inbound | Three inbound calls in a row averaging 7 overall. Close includes a day and time |
| 4 | Full calls against the cold prospect. One review per call. | 15 | Cold | A next step from the cold prospect roughly one call in five. Objection handling trending up over the week |
The numbers in the last column are ours, from watching teams run this, not an industry benchmark. The shape matters more than the figures: build the first fifteen seconds until they are automatic, add one objection at a time, and only then put the rep against the prospect who wants them gone.
Week one, in detail. The rep is not trying to book anything. They are trying to get the words out in order. Name, company, an honest acknowledgement that this is a cold call, and one specific reason for calling that fits the prospect's role. Twenty runs a day against a friendly prospect. On Friday, listen to five together and count how many times the rep said "just", "quickly" or "I know you're busy". Those are apologies, and the prospect hears them.
Alongside the dialling, the rep listens to ten recorded real calls, five that booked and five that did not, and writes one line on each. It is the fastest way to learn what a good call sounds like in your market.
Week two. Add the objection. Take the three your team hears most. On most floors that is some version of "not interested", "send me an email" and "we already have something". Run each one until the rep can acknowledge it, ask one question and get back to the conversation without looking at notes. From midweek, switch to the busy inbound prospect, who gives short answers and pushes on timing and budget. The rep will feel the difference immediately. That is the point.
Week three. Full calls, still against the inbound prospect. Two discovery questions, not eight. A cold call is a pitch that becomes a conversation, not a discovery meeting. The close is a specific ask for a short call with a proposed day and time. If the rep ends a call with "I'll send you some information", that call does not count toward the three in a row.
Week four. The cold prospect. Interrupts, asks "who is this and why are you calling", mentions the incumbent, tries to end the call at least twice. The rep will lose most of these, and should. The goal is composure, not conversion. Review every call, read the worst transcript of the day with the rep for ten minutes, and pick one line to change tomorrow. If you want a rep to feel what this level is like before you set it up, our free three-minute call against Karen is the cold prospect turned up as far as it goes.
What a manager does each week
Not much, if the format is right. Fifteen minutes on Monday as a team on one real recording and the moment the call turned. Through the week, nothing. The rep dials privately. On Friday, read each rep's best and worst scored call, ten minutes per rep, and run one twenty-minute roleplay with the rep whose score did not move. The roleplay goes where the data says, not where the calendar does.
Score every practice call the same way or the numbers mean nothing. Six skills is enough: opener, reason for the call, discovery, objection handling, value proposition, close. A number for each and one sentence that quotes what the rep said. "Objection handling, 4: when she said 'we've got something', you went to features instead of asking what it does not cover." Our page for sales managers shows how that looks across a whole team, weakest skill first, but a spreadsheet and a strict manager get you most of the way.
Two rules for the scoring. Score from the transcript, not from memory. And be stingy, or the scale tells the rep nothing.
What "ready for real dials" looks like
Not a date. A set of numbers that have moved. In our observation a rep is ready when three things are true at once: the opener and reason for call average seven or better at the inbound level, objection handling has climbed over the final week rather than bounced, and the rep has got a next step from the cold prospect at least a handful of times without a lucky prospect. A rep who hits those in week three goes live early. A rep who has not hit them by the end of week four does not go live, and that is the cheapest possible time to learn it.
When they do go live, start them on a segment where mistakes are cheap. Closed-lost from two years ago, or a territory that is fine to warm up on. Keep the practice calls going at five a day for the first month of real dials, on whichever skill the real calls show is weakest. The habit of practising is the one that outlasts the plan.
Questions people ask
- How can you practice cold calling without real prospects?
- Three ways: a colleague plays the prospect, the team reviews recordings of real calls, or the rep dials an AI simulator that plays the prospect and scores the call. Roleplay gives you a human who can stop and explain. Review gives you what real buyers said. A simulator gives you repetitions, which is the thing cold calling needs most.
- How many practice cold calls should a new SDR make before going live?
- Around a hundred over four weeks, in our observation, with the first forty spent only on the opener and the reason for the call. The number matters less than the trend. A rep whose objection-handling score is climbing after eighty calls is ready. A rep whose scores are flat needs a different kind of help.
- Is practising cold calls on an AI prospect realistic?
- It is realistic enough for the part that needs practice, the first thirty seconds and the first objection. A good simulator interrupts, pushes back and only agrees to a next step if the rep earns it. It will not replace listening to real buyers, which is why the plan below uses recorded calls alongside it.