Sales skills are the highest-leverage skill a freelancer can develop. A 10% improvement in close rate translates directly to 10% more revenue from the same pipeline, no additional marketing, no additional lead generation, no additional time.
The problem: without feedback, you don’t know what to improve. You have an instinct about how the call went, which is almost always wrong. Calls you thought went well didn’t close; calls you thought went poorly occasionally turned into your best clients. Your instinct is unreliable.
Objective feedback from call analysis changes this. The AI scoring system below gives you specific, actionable feedback after every call, not the vague “do better” that most sales advice offers, but scored dimensions with specific moments to review.
The Recording Setup
Tools:
- Otter.ai ($16/month), transcribes while you’re on the call, works with Zoom, Teams, Google Meet
- Fireflies.ai ($19/month), similar features, slightly better formatting for long calls
- Rev.ai (pay-per-use at $0.25/minute), if you don’t do high call volume
The consent protocol:
At the start of every call: “I record calls for my own note-taking. Is that okay with you?”
In practice, fewer than 5% of prospects decline. If they do, respect it and don’t record. Note that in some jurisdictions, you’re legally required to disclose recording regardless of whether you ask permission, check your local rules and always disclose.
After the call, the transcript is ready in Otter or Fireflies within 5 minutes. You don’t need to clean it up, AI handles messy transcripts well. The speaker labels (“Speaker 1” / “Speaker 2”) are enough for Claude to work with.
The Evaluation Prompt
Copy this prompt exactly and paste it with the transcript:
The Full Scoring Prompt:
“Evaluate this sales call transcript. The salesperson is [your name or ‘Speaker 1’]. The prospect is [their name or ‘Speaker 2’].
Score each dimension 1-5 (1 = needs significant work, 3 = acceptable, 5 = excellent) and provide specific improvement notes for each.
Dimension 1: Talk-time ratio Target: 40% salesperson / 60% prospect. Calculate approximate ratio from transcript. What would have been the ideal moments to ask a question instead of continuing to speak?
Dimension 2: Open question quality How many open questions were asked? Were they specific to this buyer’s context, or generic? List the 2 best questions asked and the 2 missed opportunities where a question would have served better than a statement.
Dimension 3: Objection handling Were any objections raised (price, timing, competition, internal resources)? How were they handled, addressed directly, deflected, or ignored? Provide the exact transcript excerpt and the stronger response.
Dimension 4: Next steps Was a specific next step agreed to before the call ended? If yes, was it concrete (specific date, specific action)? If no, what should have been proposed?
Dimension 5: Overall discovery effectiveness At the end of this call, did the salesperson have: (a) clear understanding of the prospect’s problem, (b) business impact of that problem, (c) decision process, (d) timeline, (e) budget range? Note which were obtained and which were missing.
Priority improvement for next call: One specific, actionable thing to change on the very next call.”
What Scores Look Like in Practice
Here’s what a scored evaluation looks like after a real call:
Talk-time ratio: 3/5 “Salesperson spoke approximately 55% of the time. The gap was most pronounced at 12:30-18:00 when you described your process for 6 minutes without asking a question. At 13:45, when the prospect said ‘we’ve tried similar things before,’ a follow-up question (‘what specifically didn’t work about that?’) would have surfaced critical information instead of continuing the service overview.”
Open question quality: 4/5 “7 open questions asked. Best questions: ‘What does success look like for you 6 months from now?’ and ‘Who else will be involved in this decision?’ Missed opportunity at 22:10, after the prospect mentioned budget pressure, a follow-up like ‘Can you tell me more about the constraint there?’ would have surfaced whether this was a real objection or a test.”
Objection handling: 2/5 “At 31:15, prospect said ‘your price is higher than we expected.’ Response was to explain value at length. More effective: acknowledge the concern, ask what they expected the investment to be, then work from that anchor. The explanation-without-acknowledgment pattern appeared twice.”
Next steps: 5/5 “Clear next step established: proposal delivery by Thursday, follow-up call scheduled for following Tuesday at 2pm. Excellent.”
Overall discovery: 3/5 “Problem: clear. Business impact: partially, revenue impact was mentioned but not quantified. Decision process: obtained. Timeline: unclear, not asked. Budget: raised by prospect but not explicitly confirmed.”
Priority improvement: “Before ending the next call, explicitly confirm the budget range. You’ve gotten the decision process and timeline down, budget is the remaining gap in your qualification.”
This feedback is specific enough to act on immediately. The scores track over time, and the patterns across multiple calls reveal your systemic strengths and weaknesses, not the random variation of individual calls.
The most common reason freelancers don’t improve their sales skills is that they never see themselves as the prospect sees them. You’re inside the call; you can’t observe yourself. The transcript + AI analysis creates the outside view you can’t generate from memory alone.
The 8-Week Improvement Cycle
Track your scores on a simple spreadsheet:
| Call date | Talk ratio | Open Q’s | Objection | Next steps | Discovery | Avg |
|---|---|---|---|---|---|---|
| Week 1 Call 1 | 3 | 3 | 2 | 4 | 3 | 3.0 |
| Week 1 Call 2 | 2 | 3 | 3 | 5 | 3 | 3.2 |
| … |
The averages across 4 calls per week give you a weekly score. The weekly scores plotted across 8 weeks show your trajectory.
Week-by-week focus areas:
Weeks 1-2: Establish baseline. Don’t try to change anything yet, just score honestly and see where you actually are.
Weeks 3-4: Focus on one dimension only. Pick the lowest-scoring dimension and address it exclusively. If talk-time ratio is 2/5, your only goal for these two weeks is asking more questions and pausing.
Weeks 5-6: Add a second focus dimension while maintaining the first improvement.
Weeks 7-8: Work all five dimensions. The improvements from weeks 3-6 should be stable; you’re now refining.
After 8 weeks, compare your Week 8 average to your Week 1 baseline. A 0.5-1.0 point improvement across dimensions is typical. More importantly, compare your close rate before and after, the skills improvement translates to actual conversion.
The Objection Transcript Library
Every time AI identifies a poorly-handled objection in your transcript, log it:
- The exact objection as the prospect stated it
- Your actual response (from transcript)
- The stronger response Claude suggested
After 20 calls, you have a personal objection library. Before every call, review the 3-5 most common objections and the stronger responses. The objections stop being surprises and start being practiced patterns.
The Talk-Time Fix
If your talk-time ratio is consistently above 50%, this single technique works: after you finish a statement, ask a question.
Every substantive statement you make can be followed with a question that invites the prospect’s reaction:
- “We’ve helped companies in your space increase lead volume by 40%. [Pause.] How does that compare to what you’re currently seeing?”
- “This engagement typically runs 3 months. [Pause.] Does that timeline align with what you’re working toward?”
- “Most clients in your situation start by addressing the content gap. [Pause.] Is that consistent with what you’ve identified as the priority?”
Each of these inverts the talk ratio for the next exchange. Over a 45-minute call, this habit shifts the ratio from 55/45 to 42/58, meaningfully into the target range without the call feeling like an interrogation.
Sales coaching used to require hiring someone who could sit in on calls, observe, and give feedback. AI makes this available after every call, at no additional cost, with feedback that’s more specific than most human coaches provide, because it’s analyzing the actual transcript, not observing from memory.
What Happens at 6 Months
Twelve weeks of scored call data tell you where your skills are. Twenty-four weeks tell you where your skills were and how far they’ve come.
Most solos who maintain this system for 6 months report two concrete outcomes:
- Close rate improvement of 10-20 percentage points
- Fewer calls needed per closed deal because qualification has improved
The second outcome is more important. It’s not just that you close more of the calls you take, it’s that you’re taking better calls. The discipline of reviewing your own performance improves how you qualify, which means fewer hours in calls with prospects who were never going to close.
The hours recovered from unwinnable calls, multiplied by your billing rate, and added to the revenue from improved close rate: this is what 6 months of AI-assisted call coaching is worth.
Ready to send stronger proposals?
Build, send, and track proposals in one place so follow-up is easier.
Start your free trial →
