How to analyse outbound call results: which numbers matter
Sado AI team · · 4 min read
In short
Read outbound call results as a funnel: numbers on the list → people who answered → real conversations → the goal (booked, confirmed, interested). Work out the share at each step and find where you lose the most. Few answers point to the hours or the list; hang-ups in the first seconds point to the opening; lots of “not interested” point to the offer or the audience. Work out the cost of one result too. Beyond the numbers, listen to 10–15 recordings.
The four key numbers
| Number | How to calculate it | What it shows |
|---|---|---|
| Answer rate | people who answered ÷ people called | Are the hours right, is the list fresh |
| Conversation rate | calls longer than 30 seconds ÷ people who answered | Do the first lines keep the customer on the phone |
| Goal rate | booked, confirmed or interested ÷ conversations | Do the offer and the brief work |
| “Don’t call” rate | people who said it ÷ people who answered | Are the list and the reason for calling right |
The 30 seconds is only an example threshold. Pick one that suits you and use the same one every time, or you can’t compare auto-calls.
Example: calling former members
Ipak Yo‘li Fitness called 800 people who hadn’t renewed their membership within a year. The goal: book a free trial session. These numbers are only an example:
- Of 800 numbers, 760 were accepted; 40 were wrong or incomplete.
- Answered: 418 of 760, or 55%.
- Had a conversation: 334 of 418, or 80%.
- Booked a trial: 50 of 334, or 15%.
- Said “don’t call”: 21 of 418, or 5%.
The biggest loss is people who didn’t answer: 342 of them. So the first fix is retries and different hours. The offer isn’t bad: 15% of the people reached booked.
What to fix when a number is low
- Few answers: change the calling time (for example, evening instead of morning), add attempts, check when the list was collected.
- Hang-ups in the first seconds: shorten the greeting and say the reason for the call at once. More: the first ten seconds of an outbound call.
- Lots of “not interested”: check the offer (is there a clear benefit?) and the list (do these people need this offer?).
- Lots of “call me later”: the time is wrong. Call those customers back at the agreed time, and start the next list at different hours.
- The “don’t call” rate went up: you’re calling the same people too often, or strangers’ numbers got into the list.
What one result cost
Divide spending by results, not by calls: total talked minutes × price per minute ÷ number of results.
Continuing the example: the 418 answered calls took 627 minutes: 627 × 800 = 501,600 so‘m. 50 bookings: 501,600 ÷ 50 ≈ 10,000 so‘m per booking. Add how many trial visitors bought a membership and you get the cost of one new member. Compare it with the cost of one lead from ads. That shows which auto-call paid off.
Where to find the results in Sado
The auto-call page shows each person’s status, attempts, result, interest, a short summary and the next step. “Download results (Excel)” puts it all in one file, where you can calculate the rates. The recording and full transcript of every call are in the portal too.
You can also ask Yordamchi, the portal’s assistant, in plain words: “How are the auto-calls going?” or “Why did this list stop?”. Once a week, listen to 10–15 recordings: the numbers show what happened, the recordings show why. More: AI outbound calls and pricing. Try it live or call 78 778 01 23.
Questions
What is a good answer rate?
There’s no general norm: it depends on the list, the hours and the reason for calling. Fresh requests usually answer more often than an old list. Compare your own auto-calls with each other.
When should we analyse the results?
After the first 50–100 calls, to fix the brief, and when the list is finished, for the full picture. Don’t draw big conclusions from small numbers.
Do unanswered calls count as spending?
No. Sado charges only for talked minutes: 800 so‘m a minute, with no monthly fee.