First response time is the least glamorous metric in customer operations and the one most tightly coupled to revenue. It is not a proxy for service quality. It is a proxy for whether the customer is still paying attention when you answer.
The research on speed to lead is consistent and brutal: responding within five minutes rather than thirty raises the odds of a meaningful conversation by roughly an order of magnitude. After an hour, most of the intent has evaporated. The customer has moved on, usually to whoever answered first.
Key takeaways
- Measure the median and the 90th percentile. The average hides your worst experiences.
- Most delay is queue time, not typing time, so hiring rarely fixes it.
- Automating the top three question types typically moves FRT more than adding two agents.
- An instant acknowledgement is not a fix. Customers grade the first useful answer.
Measure it properly first
Most teams measure this wrong in two ways.
First, they track the average. One conversation that sat over a weekend drags the mean into fiction while ten quick replies flatter it. Track the median for the typical experience and the 90th percentile for the bad one. The gap between them is where your churn lives.
Second, they count automated acknowledgements as responses. "Thanks, we'll be with you shortly" is not a first response. If your dashboard counts it, your FRT is fictional. Measure to the first message that actually addresses the question.
Segment before you optimise. Split FRT by channel and by hour. Teams usually discover their problem is not "support is slow" but "WhatsApp after 6pm is slow". That is a far more tractable problem.
Where the time actually goes
Break a slow first response into its parts and the same distribution shows up nearly everywhere:
- Queue time: the message sits unassigned. Usually 80% or more of total FRT.
- Triage time: someone reads it and decides who owns it.
- Research time: the agent looks up an order or a policy.
- Typing time: actually writing the reply. Almost always the smallest slice.
Hiring reduces typing time, which was never the bottleneck. That is why adding people so often fails to move the number.
Five changes that work
1. Automate the top three question types
Pull last month's conversations and cluster them. Almost every team finds three question types covering half of all volume, and the answer is identical every time. Order status, opening hours, and returns policy are the usual suspects. Automating those three removes half your queue, which collapses queue time for everything else too.
2. Route on content, not round-robin
Round-robin distributes evenly and ignores fit, so a billing question lands with someone who has to reassign it. Routing on what the message is about eliminates a full handoff cycle.
3. Fix the overnight gap
For most consumer businesses, a third of inbound arrives outside working hours. If those sit until morning, your 90th percentile is permanently ruined no matter how fast the day shift is. This is the single highest-leverage fix available, and it is the one automation solves outright.
4. Put context next to the conversation
If answering means opening your commerce admin in another tab, you have added a minute per conversation and an opportunity to get distracted. Order history beside the thread removes research time entirely.
5. Escalate with a summary, not a transcript
When automation hands off, the human should receive what has been established, not fourteen messages to read. This is the difference between a handoff that saves time and one that just moves the work.
What good looks like
| Channel | Customer expectation | Realistic target |
|---|---|---|
| WhatsApp / SMS | Minutes | Under 2 min median |
| Live chat | Seconds to minutes | Under 60 sec median |
| Instagram / Messenger | Under an hour | Under 15 min median |
| Hours | Under 4 hours median |
Messaging channels carry chat expectations. A two-hour reply on WhatsApp reads as being ignored, while the same delay on email is unremarkable. Adopting a messaging channel means adopting its response norms.
Frequently asked questions
Is first response time more important than resolution time?
For acquisition, yes. Speed determines whether the conversation continues. For retention, resolution time matters more. Track both and do not trade one away.
Do auto-replies improve FRT?
They improve the reported number and not the customer experience, unless the auto-reply actually answers something. Measure to the first useful answer.
What is a realistic improvement?
Teams that automate their top three question types and close the overnight gap typically cut median FRT by 70% or more, largely because queue time collapses.