Feedback said out loud, not a survey binned

Why a call rather than a message

A message survey asks the customer to stop, tap and fill in — and every one of those steps loses people. A short call only asks them to answer, and crucially lets the customer say what actually bothered them in their own words rather than collapsing it into a 1-to-5 scale. What comes back is different in kind: not a score, but a sentence you can act on — "I waited a week for the part" or "the technician was excellent but I had no idea when he was coming". The flip side is that the call has to be genuinely short, which is why the script caps the number of questions.

Catching an unhappy customer in real time

This is the biggest difference from any other feedback method. When a customer says something negative, the agent does not move on to the next question as if nothing happened — the script defines that node, and it can bridge a service manager into the same call immediately, extension dialling included. If nobody can take it, the call is flagged as a priority escalation with the reason, so it sits at the top of the morning list. The odds of saving a frustrated customer while they are still on the line are far better than after they have written a review.

What comes back to you

Every call has a summary, sentiment, outcome, and the captured answers as fields — alongside the full transcript, where the quotes live. On the systems side, a summarised activity is written into Salesforce or HubSpot and a signed webhook fires, so the feedback sits on the customer record rather than in a separate sheet. Identifying details in the transcript are masked automatically, and a deletion request runs through the full anonymisation flow — particularly relevant when the feedback concerns a sensitive service.

• Sentiment and a summary per call, aggregatable over time
• Structured answers as fields, open answers as quotes in the transcript
• A CRM activity and a signed webhook at the end
• PII masking and full anonymisation on request

Timing: when to call, and who not to call

Timing matters more than wording. Too close to the event and the customer is still mid-process; too far and they no longer remember. The practical rule is to call once the service has genuinely finished and within reasonable hours, not in the evening or at weekends. Equally important is who not to call: a customer with an open complaint, a customer who asked not to be contacted, and a customer surveyed recently. Those rules are set on the list and the campaign — a refusal is recorded and honoured, and a lead marked do-not-contact leaves the queue.

The limits, unvarnished

A feedback call is an outbound contact, so the same rules apply as to any outbound dialling: responsibility for consent, suppression lists and calling hours sits with the business, and the system supplies the record and a do-not-contact flag — not a policy. Three practical limits: not everyone will answer, and that is fine as long as you do not try five times; if the service itself has not improved, repeated feedback calls will wear customers down rather than please them; and in sensitive services — medical, therapeutic, legal — think hard about whether an outbound call is the right channel, and at minimum allow an immediate opt-out in the opening.