AI sales call scoring with written reasons.
After every call Saley gives you an overall 0–100 score, a five-dimension rubric with written reasons, and a BANT checklist. When Saley returns supporting BANT evidence, you can compare the transcript line with what was actually said.
The moment you stop
An immediate call snapshot while the full review finishes.
Live Assist turns into a recap as soon as you end the call. It shows the best available summary, up to three takeaways, and detected action items, then the structured notes and coaching enrich the review when they are ready.
The score
One number, five dimensions.
The overall 0–100 is a quick read on the whole call. Underneath it sits a rubric of five dimensions sellers already think in — so a low score tells you where it went sideways, not just that it did.
A practice call is scored the same way, so a drill with a colleague is directly comparable with a customer call. AI sales roleplay and call practice
Grading calls by hand as well? The same five dimensions, a three-point scale, and a box for the quote behind every score: sales call scorecard template
- Discovery. Did you surface the real problem, or pitch past it?
- Qualification. Is this a fit worth your time — and did you find out?
- Value delivery. Did the pitch connect to what they actually said hurt?
- Objection handling. Did you meet the pushback, or route around it?
- Next steps. Did the call end with a committed next move?
Scored against your playbook
- Discovery90
- Qualification74
- Value delivery85
- Objection handling80
- Next steps82
- BudgetNot raised on the call
- Authority“I'd loop in my VP”
- Need“notes are a mess”
- Timeline“next quarter”
BANT coverage
See what was covered and the evidence Saley found.
Budget, Authority, Need, Timeline — each one is marked covered or not. When the analysis returns supporting evidence, the BANT item shows the relevant line from the transcript.
A score needs an explanation. The rubric gives you written reasons, while BANT items show supporting transcript lines when available. You can compare the result with the call and correct it when the model gets something wrong.
Graded on your terms
Scored against your Playbook, not a generic ideal.
The same Playbook that powers the live overlay grades the call — your value props, your objection responses, your call goal. So “good discovery” means good discovery for what you sell, and the advice is specific enough to act on next time.
When it runs
Runs after the call ends.
The review runs when the call is over, from the stored transcript (on very long calls, its later part). If it fails, Saley retries once by itself, and you can regenerate it up to twice per call without changing your notes, follow-ups or current focus. It complements the live overlay rather than replacing it: help in the moment, then a clear-eyed grade afterward.
Notes vs the score
What the call was about, and how you did — two different things.
Alongside the score, Saley writes auto call notes — a running recap during the call and a structured summary after: topics, key points, decisions, action items, next steps.
The notes are descriptive — what happened, in plain terms, editable and yours. The score is the judgment — how it went against your Playbook. Keeping them apart means your recap for the CRM never gets tangled up with the grade.
The same review also lists what each side promised, with the quote and any spoken due date, and drafts a follow-up email from those promises. Sales call follow-ups
Intro call with an eight-rep sales team; pricing, onboarding and a security review discussed; next step booked.
- Pricing
- Onboarding
- Security review
- Team of eight reps, three new hires ramping next quarter.
- Reviewing calls the day after today — wants help in the moment.
- Send the security overview
- Book the next call with the VP
Coaching focus
One habit for your next call, checked against what you said.
After each call the review suggests one behaviour of 12 words or fewer, such as “Ask for a concrete next step before closing”, plus a one-line tip on how. The suggestion names its source in this call’s feedback: a weakness, a missed opportunity or a rubric dimension.
- On the next call it shows as one quiet line in the overlay. The next review marks it done, partly or not observed.
- Done and partly require a line you said, quoted word for word. A done result shows the time and the quote; click it to jump to that moment.
- You can accept it, keep your current focus, reword it, write your own, or clear it.
- After two done calls in a row Saley offers a new focus; it never switches on its own.
- Calls can be filtered by focus result.
Calls under two minutes, or with fewer than five lines from you, aren’t graded and don’t set a focus. A paraphrase doesn’t count. On very long calls the review reads the later part of the transcript, so something you did early can come back as not observed. Results aren’t charted across calls: each call shows its own result.
AI call scoring FAQ
What the post-call review measures
How does Saley score a sales call?
Saley produces an overall score from 0 to 100 and reviews five dimensions: discovery, qualification, value delivery, objection handling, and next steps. Each dimension includes a written reason rather than only a number.
Does the AI call score use my sales methodology?
The review uses the active Saley Playbook, including your value propositions, objection responses, and call goal. The fixed five-dimension rubric stays comparable between calls while the written feedback is grounded in what you sell.
Does Saley show evidence for its BANT assessment?
When the analysis returns supporting evidence, Budget, Authority, Need, and Timeline items include the relevant transcript line. You can compare the evidence with the call and re-run the review later on a stored call.
Are AI call notes and AI sales coaching the same thing?
No. Call notes describe topics, decisions, action items, and next steps. The coaching score evaluates how the seller performed against the rubric and Playbook, so the factual recap remains separate from the performance judgment.
Does Saley tell me what to work on before my next call?
Yes. Each review suggests one behaviour to practise, shows it as a single line in the overlay on your next call, and then marks it done, partly or not observed, quoting your own words when you did it. Calls under two minutes aren't graded.
Keep reading
The rest of the loop
Get Saley
See the overlay on your own calls.
Start with the scripted walkthrough, then put the live overlay on your own calls. From $69/month.
No free tier · No trial · Billed today · Cancel anytime · macOS 13+ · Apple Silicon & Intel