Talk-to-Listen Ratio Calculator
A paste-in method for calculating your talk-to-listen ratio from any call transcript, plus the benchmark data on what top discovery performers actually do differently: no software required, just a transcript and five minutes.
How to use it
Paste your call transcript into the speaker-tagged format described, run the word-count formula manually or via the provided spreadsheet formula, and compare your ratio against the benchmark table, track the trend across multiple calls rather than judging off one.
What's inside
- The core formula (word-count and time-based methods)
- Step-by-step manual calculation walkthrough
- Copy-paste Google Sheets / Excel formulas
- Speaker-tagging rules for messy transcripts
- Benchmark table: ratio by performance tier
- A worked example using a sample transcript
- Scoring interpretation bands
- Guardrails: what the ratio does and doesn't tell you
The Core Formula
Talk-to-Listen Ratio = (Rep words ÷ Total words): (Prospect words ÷ Total words), expressed as Rep:Prospect (e.g. 43:57).
Two ways to measure it:
| Method | How | Accuracy |
|---|---|---|
| Word-count method | Count words per speaker in a text transcript | Good, fast, works from any transcript |
| Time-based method | Sum seconds each speaker talks using timestamps | More accurate, accounts for pace and pauses |
Step-by-Step: Word-Count Method (Manual)
- Get your transcript with speaker labels (e.g. "Rep:" and "Prospect:" before each line). Most call-recording tools export this automatically; if not, tag each line yourself.
- Copy all Rep lines into one block. Count the words (most word processors show a live word count).
- Copy all Prospect lines into another block. Count the words.
- Total words = Rep words + Prospect words.
- Rep % = Rep words ÷ Total words × 100. Prospect % = 100 − Rep %.
- Express as Rep:Prospect (e.g. 48:52).
Copy-Paste Spreadsheet Formula (Google Sheets / Excel)
Set up two columns: A = Speaker ("Rep" or "Prospect"), B = Line Text, one row per line of transcript.
In column C, next to every row, paste: `` =IF(A2="Rep", LEN(TRIM(B2))-LEN(SUBSTITUTE(TRIM(B2)," ",""))+1, 0) `` This counts words in that line if it's a Rep line, else returns 0.
In column D, paste the mirror formula for Prospect: `` =IF(A2="Prospect", LEN(TRIM(B2))-LEN(SUBSTITUTE(TRIM(B2)," ",""))+1, 0) ``
Then at the bottom: `` Rep total words: =SUM(C:C) Prospect total words: =SUM(D:D) Rep %: =SUM(C:C)/(SUM(C:C)+SUM(D:D))*100 Prospect %: =100-Rep% ``
Speaker-Tagging Rules for Messy Transcripts
- If your export has no speaker labels, tag lines manually using audio playback, don't guess from context alone.
- Cross-talk / interruptions: assign the words to whoever completes the thought, not whoever started the interruption.
- Filler-only lines ("mm-hmm," "right," "yeah") from the listener count as words but should be sanity-checked: if they're inflating one side's total, note it separately rather than let it skew the read.
- Exclude pre-call small talk and post-call logistics from the count, measure the discovery portion of the call only.
Worked Example
`` Rep: So walk me through how this is handled today. Prospect: Right now it's basically three people manually reconciling spreadsheets every Friday, and it usually runs into the weekend. Rep: What happens when it runs late? Prospect: Finance can't close the month on time, and that's become a standing agenda item with my VP. ``
Rep words: 17. Prospect words: 39. Total: 56. Rep % = 17/56 × 100 = 30%. Prospect % = 70%. Ratio = 30:70: rep-light in this snippet, which is a healthy sign this early in discovery.
Benchmark Table
| Ratio (Rep:Prospect) | What it usually means |
|---|---|
| 70:30 or worse | Pitching, not discovering: rep is dominating airtime |
| 55:45 to 65:35 | Still rep-heavy; room to ask more and explain less |
| 43:57 to 50:50 | Commonly cited benchmark range for top discovery performers (widely referenced call-analytics research, e.g. Gong's studies on winning discovery calls) |
| 35:65 to 42:58 | Strong listening ratio: verify the rep is still setting agenda and asking sharp questions, not just going quiet |
| Below 35:65 | May indicate lack of control or agenda-setting rather than good discovery, investigate the transcript directly |
Scoring Interpretation
- In benchmark range (43:57–50:50): trend this call as a model example for the rep's own reference.
- Rep-heavy (55:45+): flag for coaching, check whether the call slipped into pitching before pain was fully surfaced.
- Extremely prospect-heavy (below 35:65): don't assume this is automatically good, check that the rep still asked enough questions to actually qualify the deal.
Guardrails: What This Ratio Does NOT Tell You
- It says nothing about question quality: a rep can hit 45:55 asking only closed, surface-level questions.
- It doesn't account for necessary rep talk time like teaching moments or reframing a prospect's assumption, some rep talk is high-value.
- Use it as a trend metric across multiple calls, not a single-call verdict. One good or bad ratio on one call proves little; a pattern across 10+ calls is meaningful.