Indie call coaching · No enterprise contract · Built for the AE who closes their own loop

Score every call.

Drop a sales call transcript. Get a 14-point teardown in 90 seconds. The bell rings when something breaks.

Try the demo → How it works
14.
→ Points scored per call
90saverage score time
$19/movs Gong's $1,500/seat
What gets scored ↓
→ The rubric · 14 dimensions

What gets scored.

Same rubric every call. No "make it punchier" generic notes. Each dimension has a numeric output and a why.

Built from the patterns we saw in 1,200 anonymized B2B sales-call transcripts. Tuned against deal-won and deal-lost outcomes. The 14 points predict close-rate within ±8% on calls we've back-tested.

01 Talk-listen ratioTarget: 43:57 prospect-to-rep on discovery 10%
02 Question densityOpen-vs-closed mix · Discovery vs. confirmation 9%
03 Discovery depthBANT / MEDDIC coverage · Pain-metric-decision-process 12%
04 Pain articulationDid the buyer describe the problem in their own words? 9%
05 Buying-signal mentions"When could we start" · "what's the cost" · "who else uses this" 8%
06 Objection handlingQuality · Bypass detection · Reframing 8%
07 Next-step claritySpecific date · Specific people · Specific action 8%
08 Pitch-vs-discovery timeTime on selling vs. time on understanding 7%
09 Filler-word densityUm · uh · like · you-know · per minute 5%
10 PacingTalk-bursts vs. silences · Strategic-pause detection 5%
11 Decision-maker confirmationDid the rep confirm authority before pitching? 7%
12 Champion-language detection"We," "our team," "I'll bring this to my boss" 6%
13 Risk flagsSingle-thread · "Let me check budget" · "Not now" 3%
14 Composite closer scoreWeighted blend · 0–14 · Trends across calls 13%
∑ Closer ScoreOut of 14 · The number that matters 14.0
→ Try it · Live engine · Sample transcripts pre-loaded

Score a real call.

Three samples below — pick one, hit Score, watch the ticket print. Or sign in and paste your own for a real score in 90 seconds.

Sample transcripts
Pick a sample on the left.
Hit "Score this call".
The ticket will print here.
How it works ↓
→ Four steps · No setup · No CRM connect

What you do.

01

Get the transcript

Paste from Fireflies, Gong, Otter, Krisp, or Zoom's native transcription. We accept any text format.

02

Hit Score

Drop the transcript into Closerbell. The 14-point engine runs in 60–90 seconds. No setup, no calibration call.

03

Read the ticket

Composite score · standout move · risk flag · the one coaching action to try next call. Print-ready.

04

Score the next one

Trends compound after 8 calls. Compare won-vs-lost. Forward to your manager when something interesting fires.

→ Pricing · No annual contract · Cancel any time

Score for free. Or scale.

Free tier hooks the curious. $19/mo for the daily AE. $99/mo bundles it with two other Savy-stack tools. $299+/mo for sales managers running team scoring.

Free

Drop in.

$0/mo
  • 1 transcript / day
  • 14-point teardown
  • One coaching move
  • Session-only (no save)
Try it →
Most popular

Closerbell.

$19/mo
  • 50 transcripts/mo
  • Save + revisit scores
  • Closer-score trend chart
  • Compare won-vs-lost
  • Custom 14-point rubric
  • Weekly email digest
Score a call →
Bundled

SD Starter.

$99/mo
  • 500 transcripts/mo
  • + Savy domain audits
  • + MTN Verify lists
  • Cross-tool dashboard
  • Email + chat support
Score a call →
For managers

SD Pro+.

$299/mo
  • Unlimited transcripts
  • Team scoring view
  • Per-AE scorecards
  • SDCallix auto-capture
  • Priority support
Score a call →
→ Public receipts wall · Coming with launch

We score famous calls.

Every Friday we post one anonymized scoring of a publicly-known sales-call clip. The pattern shows up after about 6 of them. Send us a clip you want scored.

See the wall →
12.5
SaaS Pivot Pitch
discovery depth: 9/10
standout: re-framed at 18:42
8.5
Series-A Demo
talk-listen: 71:29 ⚠
risk flag: budget evasion
11.0
Agency Discovery
question count: 12 (8 open)
standout: silence at 14:22
6.0
Cold Outbound
pain articulation: weak
risk flag: single-threaded
→ Asked five times in week one

The questions.

Do you train on my transcripts?
No. Transcripts are scored, the score is saved to your account, the raw text is dropped after 30 days unless you explicitly tag it as a "saved example." We don't fine-tune any model on your data. Ever.
How is this different from Gong / Chorus?
Gong and Chorus are enterprise tools — $1,500+/seat/year, multi-month implementation, designed for sales managers who want a forecasting view across a 50-rep org. Closerbell is the indie AE / SDR self-serve tool — $19/mo, no setup, designed for the rep who wants to coach themselves between manager 1:1s. Different ICP, different price, different problem.
Can I upload audio?
Not at v1. Transcript only. Most call platforms (Fireflies, Gong, Otter, Krisp, Zoom) generate transcripts automatically — paste from there. Audio support comes when we have a clear answer on storage costs and privacy compliance.
What if my company has compliance rules about call data?
If your company forbids pasting customer-call data into third-party tools, Closerbell isn't a fit until SD Pro+ lands (which includes a self-hosted scoring option). For personal coaching with your own internal call records — Closerbell is fine. For regulated industries (healthcare, finance) — talk to us first.
How accurate is the score?
The 14 dimensions are tuned against deal-won and deal-lost outcomes from a back-test of 1,200 anonymized B2B sales transcripts. The composite closer-score predicts close-rate within ±8% on the holdout set. We publish the validation methodology in our field notes.
Is the engine actually live?
Yes. Sign in, paste a transcript, get a 14-point teardown in 90 seconds. Free tier is 1 score per IP per 24h. No credit card.

Score your first call.

Magic-link sign-in. Paste a transcript. Get a 14-point teardown in 90 seconds. Free tier is one score every 24 hours. No credit card.

Score a call → See the samples first
Live engine · Gemini Flash + Cerebras + Groq fallback · 90s avg · Your transcript dropped after 30 days unless saved.