How to hire a cold caller who actually converts: the screening guide

The mistake almost everyone makes

Businesses hiring a freelance cold caller optimise for hourly rate. It's the wrong variable, and the arithmetic shows why quickly.

A caller at $6/hr who books two qualified appointments a week costs you $3 per hour of your own time wasted on the unqualified ones, plus the opportunity cost of a market you've now partially burned. A caller at $18/hr who books eight costs three times as much and produces four times the output. The cost per qualified appointment — the only number that matters — is roughly $120 for the first and $90 for the second, and that's before accounting for show rate, which is usually far worse for the cheap caller.

The rate isn't the cost. The cost per qualified appointment is the cost, and you can't know it from a profile.

Step 1: write a job post that filters

Vague posts attract vague applicants and generate fifty proposals you have to read. Specific posts get eight, of which three are worth talking to. Include:

• The exact target market. "Owner-operators and small carriers with 1–10 trucks in the US" not "B2B leads".
• The decision-maker title you need reached.
• Your CRM. HubSpot, Salesforce, GHL or none. This alone filters heavily.
• Expected daily dial volume and hours per week.
• What the outcome is — appointments, transfers, qualified leads, or closes. These are different jobs.
• Whether the list exists or needs building.
• Your rate range. Publishing it saves everybody a round of messages.
• One screening instruction — "start your proposal with the word 'dialtone'". Roughly two-thirds of applicants won't, and you can discard them unread.

Step 2: ask for a recorded sample call

This is the single highest-signal step in the whole process and most people skip it.

Any experienced cold caller can produce a recording — of a genuine call where permitted, or a recorded mock call against a scenario you give them. Ask for both if you can: the mock tells you about their craft, the real one tells you what they're like when it isn't going well.

What to listen for:

• The first fifteen seconds. Do they earn the right to continue, or do they launch into a pitch?
• Pace. Nervous callers rush. Rushing signals a script and triggers the brush-off reflex.
• What happens after "not interested". Do they acknowledge and pivot, or do they steamroll?
• Do they ask questions or make statements? Questions keep calls alive.
• Do they listen? Listen for whether their next line responds to what was actually said or to what the script expected.
• Accent and clarity for your market. Judge this honestly and unsentimentally, because your prospects will.

Five recordings will tell you more than five interviews. If someone won't provide one, that's your answer.

Step 3: five questions worth asking

Skip "tell me about yourself". Ask these:

1. "Walk me through how you'd build the list for this campaign." Tests whether they think about data or only about talking. The ones who only talk are the ones who fail in week three.
2. "What's your dial-to-conversation rate in a market like mine, and what's the appointment rate off that?" Anyone quoting suspiciously high numbers is either measuring differently or making it up. Ask them to define each term.
3. "Tell me about a campaign that didn't work and what you concluded." Experienced callers have several. People who've never had a failure haven't run enough campaigns to be useful.
4. "How would you handle [your market's hardest objection]?" Give them a real one. You're testing thinking, not memorisation.
5. "What would you need from me in week one?" Good answers ask about the ICP, the offer, existing objection patterns and CRM access. Bad answers say "just send me the list."

Step 4: structure the paid trial properly

Never hire long on an interview. Run a 7–14 day paid trial with KPIs agreed in writing beforehand. Free trials are a bad idea — they attract people with nothing better to do and produce work nobody values.

Define these before the first dial:

• Dials per day. 120–180 is a realistic full-time range with a dialer and a clean list.
• Connect rate. Varies hugely by market. Establish a baseline in week one rather than imposing a target.
• Conversations per day. Define what counts as a conversation.
• Appointments set, against a written definition of "qualified".
• Show rate — the one that actually matters, though a two-week trial may not produce enough data to judge it.
• CRM hygiene. Every call dispositioned, every note usable by someone else.

Then be realistic about what a trial can prove. Two weeks tells you whether someone is competent, diligent and honest. It doesn't tell you whether the campaign will work — that takes six weeks, because follow-up sequences from weeks one to three are where a lot of the revenue lands.

Step 5: the four signals that reveal a bad hire early

Dispositions are all one category. If everything is logged as "no answer", they're either not calling or not logging. Either is disqualifying and both are visible in the CRM on day three.

Notes are unusable. "Interested, call back" tells you nothing. Good notes capture what the prospect actually said, in their words, and the specific objection raised.

Appointments don't show. A show rate under about 40% after the first fortnight usually means they're booking anyone who says "maybe" rather than qualifying.

They never report a problem. Every campaign has friction in week one — dead numbers, an opener that isn't landing, a target list that's wrong. A caller who reports nothing but good news is either not paying attention or not telling you the truth.

Where to hire

Upwork. The most practical route for a first engagement. Escrow, hourly tracking with screenshots, a work diary you can audit, and dispute resolution. Filter on hours logged in relevant work rather than on overall rating, which is easily inflated. Look for someone with a long-running contract in their history — that's the strongest signal available on the platform, because clients don't keep bad callers for a year.

Referrals. Highest quality, lowest availability. Ask other businesses in your space who they use.

Agencies. Faster to scale and more expensive, and you generally don't control who dials. Worth it when you need thousands of dials a week immediately; wasteful when you need one reliable seat.

LinkedIn. Works but slower, and you carry all the payment and verification risk yourself.

What to expect once you've hired

• Week 1: list building, CRM setup, opener drafted and first dials. Expect noise, few appointments, and at least one thing that turns out to be wrong about the target.
• Week 2: first appointments, opener rewritten at least once, list quality becoming measurable.
• Weeks 3–4: a rhythm emerges. Follow-up from week one starts producing.
• Weeks 5–8: this is where campaigns become predictable, and typically where output roughly doubles against the first month.

Judging an outbound campaign at four weeks is judging it at close to its worst point. If the activity metrics are right and the notes show real conversations happening, give it eight before you draw conclusions.

Dip Palit

B2B SDR · MCA & FMCSA/DOT compliance specialist

I run outbound campaigns for US businesses, with deep specialisation in merchant cash advance and motor carrier compliance. 1400+ tracked Upwork hours and $60K+ generated for a single transportation services client. More about me or get in touch.