Promises you can keep
Separate first response from resolution
First response time measures how long someone waits to know they have been heard. Resolution time measures how long they wait for the problem to end. Customers experience these differently, and collapsing them into one target hides which of the two is actually failing.
A fast first response with a slow resolution produces a specific complaint: the customer feels handled rather than helped. The fix is not a faster acknowledgement, it is a realistic commitment about what happens next and who owns it.
Set a target per channel, not one number
Expectations differ by the channel the customer chose. Live chat implies minutes; email implies hours; a form on a contact page implies a working day. Publishing one figure across all of them guarantees you look slow somewhere and over-promise somewhere else.
Set the target from observed behaviour, not aspiration. Look at your current distribution per channel, take a percentile you already hit reliably, and improve from there deliberately.
- Live chat: measured in minutes, staffed to concurrency rather than volume.
- Email and forms: measured in working hours, with the working calendar stated.
- Social and marketplaces: governed by the platform’s own expectations.
- Phone: measured in seconds to answer and in abandoned-call rate.
Staff the target you published
A target is a staffing statement wearing different clothes. Two-hour email responses across twelve hours a day, seven days a week, is a rota question long before it is a performance question, and no amount of coaching closes a gap that is really a headcount gap.
Model the target against your demand curve rather than your daily total. Support volume arrives in peaks, and a team sized for the daily average will miss the target every morning and idle every evening.
Protect the target when volume spikes
Decide in advance what gives way under pressure, because something will. Agree which queues degrade gracefully, which contact reasons get a holding response, and at what threshold you change the promise publicly rather than quietly missing it.
A published, temporary change in expectation costs far less goodwill than silence. Customers forgive a stated delay; they rarely forgive a commitment that simply stopped being true.
Report the misses, not just the average
Averages conceal the cases that damage trust. A mean of four hours can contain a tail of conversations waiting three days, and it is the tail that generates complaints, escalations and churn.
Report the percentage of contacts answered inside the target and look directly at the slowest few percent every week. That group is where your process problems live, and it is a far shorter list than it appears.