First Response Time (FRT)

Table of Contents

  1. Definition
    1. How it relates to marketing
  2. How to Calculate
  3. How to Utilize
  4. Comparison to Related Metrics
  5. Best Practices
  6. Future Trends
  7. FAQs
  8. Related Terms
  9. Sources

Definition

First Response Time (FRT) measures how long a customer waits between reaching out and getting the first real reply from your team. The clock starts when the ticket, chat, email, or message lands. It stops at the first meaningful response — one that actually begins to address the issue.

That word “meaningful” is doing real work. An automated “we’ve received your message” doesn’t stop the clock. Neither does a chatbot’s canned holding reply, unless your policy explicitly counts bot responses on self-service channels. FRT is about the moment a customer stops wondering whether their message vanished into a void.

It goes by a few other names — first reply time, time to first response — and it’s one of the most widely tracked support metrics going. Front’s research found that 47% of B2B companies measure it, which puts it near the top of the list.

One clarification that trips people up: FRT is not the same as average reply time. Average reply time covers every back-and-forth message in a conversation. FRT only counts the first one. And it’s a long way from Time-to-Resolution, which measures the whole arc through to the problem being solved. FRT is the hello. TTR is the goodbye.

Back to Top

How it relates to marketing

The first reply sets the tone for everything after it. A fast one signals that a customer’s time matters before you’ve solved anything; a slow one breeds frustration before you’ve even started. That first impression bleeds directly into brand perception, and that’s where FRT stops being a back-office stat and starts being a marketing concern.

The retention math is stark. Companies that reply to emails within an hour hold onto 71% of customers, while those that take a full day drop to 48%, per figures compiled in Ringly.io’s benchmark work. Speed also shapes conversion. First-response time matters most on pre-sales questions — “will this arrive before Christmas?”, “which size fits?” — where a shopper is one slow answer away from closing the tab and buying it on Amazon instead. Gorgias frames those pre-purchase moments as some of the highest-leverage FRT touchpoints there are.

Customers have made their expectations loud. Around 90% of U.S. customers call an immediate response “important” or “very important,” and 60% want help within 10 minutes. Miss that window and you’re disappointing more than half your audience before an agent has typed a word.

Back to Top

How to Calculate

FRT for a single ticket is just subtraction: the timestamp of the first meaningful reply minus the timestamp when the ticket was created. Averaged across a period, the formula is:

Average First Response Time = Total of All First Response Times / Number of Tickets

A worked example from Gorgias makes it concrete: 85,000 seconds of combined first-response time across 900 tickets works out to about 94 seconds average FRT. Under a minute and a half.

Two calculation choices matter more than the formula:

  • Business hours, not calendar hours. Unless you run 24/7 support, measure within your stated hours. A ticket that arrives 10 minutes before close and gets answered 15 minutes after open the next morning should read as 25 minutes, not overnight. Otherwise nights and weekends wreck the average and tell you nothing useful.
  • Median over mean. A handful of tickets that sat for days will drag the average up and make a healthy team look slow. Most B2B support teams report the median instead, and serious operations track p75 and p90 too, to see the tail — the worst-case waits some customers actually experience.

Exclude automated acknowledgments and spam or test tickets from the count, and normalize timestamps to a single time zone before you do anything else. Tools like Zendesk, Help Scout, and Gorgias calculate FRT automatically, which saves the spreadsheet gymnastics but doesn’t save you from defining the rules clearly first.

Back to Top

How to Utilize

FRT is most useful as a leading indicator — an early read on service quality that moves before CSAT and churn do. The number on its own is thin. Sliced up, it points straight at bottlenecks.

Segment by channel. Expectations swing wildly depending on where the customer showed up:

  • Live chat: seconds, not minutes. Zendesk treats around 40 seconds as strong performance; the industry average sits closer to 2 minutes.
  • Email: Zendesk’s tiers put under 12 hours as acceptable, under 4 hours as good, under 1 hour as best-in-class. The cross-industry average is a rough 7–12 hours, well short of the 4-hour mark that roughly 46% of customers expect.
  • Social media: under an hour, since a slow public reply is a slow public reply.
  • Phone: usually tracked as Average Speed of Answer under the old 80/20 rule — 80% of calls answered within 20 seconds.
  • Messaging apps (WhatsApp, Messenger): under 5 minutes; the channel itself implies speed.

Tie it to SLAs. Many contracts bake an FRT commitment into the service level agreement, especially in B2B. FRT is how you prove attainment — or spot the queues where you’re breaching before the customer complains.

Diagnose the queue, not the agent. This is the shift that changes outcomes. Slow FRT is rarely about people typing too slowly; it’s about how work gets routed and owned. Front’s Coordination Tax research found the typical B2B team spends nearly three hours on coordination for every hour actually solving customer problems. Customers wait because tickets sit in a shared inbox with no clear owner, not because the issue is hard.

Pair it with a quality check. A fast, useless first reply is worse than a slightly slower helpful one — it just generates a second ticket. Track FRT next to First Contact Resolution and an agent quality score so speed doesn’t quietly cannibalize substance.

Back to Top

FRT gets lumped in with several timing metrics that measure genuinely different things. Here’s the separation:

MetricWhat it measuresWhat triggers the “stop”
First Response Time (FRT)Time from contact to the first meaningful replyThe first substantive human (or counted bot) response
Time-to-Resolution (TTR)Total time from contact to the issue being fully solvedThe case closes, resolved
Average Reply TimeAverage wait across every reply in a conversation, not just the firstEach subsequent response, averaged
Average Speed of Answer (ASA)Time a phone caller waits in queue before an agent picks upThe call is answered
Average Handle Time (AHT)Time an agent actively spends on an interactionWrap-up/after-call work is done

The quick mental model: FRT is how long until someone says something back. Average Reply Time is how long the customer waits each time after that. ASA is the phone-line version of FRT. AHT is the agent’s hands-on time. And TTR is the only one that cares whether the problem actually got fixed.

FRT is also easy to confuse with Mean Time to Acknowledge (MTTA) from IT incident management — both mark an initial reaction — but MTTA tracks a system alert being acknowledged by an on-call team, not a customer getting a reply.

Back to Top

Best Practices

Fix routing before you push for speed. Because FRT is mostly a queue-and-ownership problem, intelligent triage and automated routing move it far more than telling agents to hurry. Lorikeet’s data suggests AI-assisted triage and resolution produce improvements several times larger than anything you’ll get from agent-side optimization alone. Assign billing questions to the billing team and API questions to engineering automatically, and the first response lands faster because nobody’s deciding who owns it.

Use automation for the acknowledgment, humans for the answer. An instant, honest “we’ve got this, here’s your ticket number” buys goodwill even when a full reply takes longer — as long as you’re not counting that auto-reply as your FRT. Bots handle initial replies dramatically faster than people and can absorb volume spikes without making anyone wait.

Lean on self-service to thin the queue. A strong knowledge base does double duty: customers self-resolve simple questions, and agents get ready-made first replies for common issues. Fewer tickets in the queue means faster first responses on the ones that remain.

Find the Goldilocks target. Geckoboard’s phrase is a good one — set an FRT goal that’s quick enough to satisfy customers but not so aggressive it stresses agents into sloppy replies. A target that torches response quality isn’t a win.

Review and adjust on a schedule. FRT drifts with volume, staffing, promotions, and product releases. Overlay ticket arrival rates and staffing levels on your FRT trend so you can explain a spike instead of just reacting to it.

Back to Top

AI is the dominant force reshaping FRT, and for this metric specifically the effect is close to total. AI-handled tickets effectively have zero queue time — the response is instant — which is why blending them into a single average can flatter the number and hide how humans are doing on the tickets that reach them. Expect more teams to report AI-handled and human-handled FRT separately.

The headline figures are large enough to warrant a second look. Third-party write-ups of Freshworks’ benchmark data cite AI cutting average first response time from over six hours to under four minutes. LivePerson’s conversational AI benchmark reports bots handling initial replies far faster than human agents. Treat the specific numbers as directional until you’ve reproduced them in your own environment — they come through vendor reporting and vary hugely with ticket mix.

There’s also a quieter shift in what FRT even means. As Genesys and others note, enterprises are moving past pure first-response speed toward first-response resolution — making sure the fast reply also solves the thing, so speed doesn’t just defer the work. In an agentic customer-experience model, where systems complete tasks rather than just answer, the interesting question stops being “how fast did we reply?” and becomes “how fast did we finish?” FRT won’t disappear, but it’s increasingly reported alongside resolution and containment metrics rather than on its own.

Back to Top

FAQs

What counts as a “response” for FRT? A meaningful, substantive reply that begins to address the customer’s issue. Automated “message received” acknowledgments and chatbot holding messages don’t count, unless your team has a documented policy of counting bot replies on self-service channels.

How is FRT different from Time-to-Resolution? FRT measures how long until the customer hears back the first time. TTR measures how long until the problem is actually solved. You can post an excellent FRT and a poor TTR at once — fast to acknowledge, slow to resolve.

Should I measure in business hours or calendar hours? Business hours, if you don’t offer 24/7 coverage. Counting overnight and weekend hours against tickets that arrived after close inflates your average and hides your real daytime performance. Many teams report both, but business-hours FRT is the one that reflects the experience.

Mean or median? Median gives the fairer read for most teams, because a few multi-day outliers can badly skew the mean. Track p90 separately to understand the worst waits some customers are actually getting.

What’s a good FRT benchmark? It’s entirely channel-dependent. Live chat is measured in seconds (around 40 for strong performance), email in hours (under 1 hour is best-in-class, under 4 is good), social media under an hour. Borrowing a single cross-channel number won’t tell you much — segment first.

Why is our FRT high even though our agents work hard? Usually coordination, not effort. Tickets sitting in a shared inbox with no clear owner, manual triage, and messy handoffs create most of the delay. Clarifying ownership and automating routing tends to move the number more than anything agents can do individually.

Does faster FRT actually improve satisfaction? Up to a point, and only if quality holds. Speed correlates strongly with CSAT, but a rushed, unhelpful first reply that generates follow-ups can lower satisfaction even as the FRT number improves. Pair it with First Contact Resolution.

Can AI reduce FRT without hurting quality? It can, especially for common, repeatable questions where an AI agent answers instantly and routes complex cases to a human with context attached. The risk is a fast but irrelevant automated reply, which frustrates more than it helps — so quality monitoring matters as much as speed.

Back to Top
Back to Top

Sources

Back to Top

Was this helpful?

Previous

Employee Lifetime Value (ELV)

Next

Forrester Total Experience Score