If you run customer service for an online store or a service company, response time is the first thing customers judge you on, before the reply is even read. This article goes through what the latest industry reports say about customer expectations in 2026, channel by channel, and why "first reply" and "resolved" are two different measures. The material is international surveys; Swedish figures are thin, and where they are missing we say so.
How fast a response do customers expect in 2026?
Customers expect faster replies than a year ago, and what they really want is a resolution, not a reply. According to Zendesk's CX Trends 2026, based on responses from over 11,000 consumers and businesses in 22 countries, 88 percent of customers expect faster replies than a year ago, and 74 percent expect customer service to be available around the clock.
The expected time to resolve the issue has shrunk too. In HubSpot's State of Service, more than half of the managers surveyed say customers now expect their issue resolved within three hours or less. HubSpot's compilation of customer service statistics also reports that 90 percent of customers consider an immediate response important or very important when they have a question, and that 60 percent define immediate as ten minutes or less.
The figures are self-reported expectations, not measured behaviour, and they come from vendors who sell tools for faster replies. But the direction is the same in every report, and it matches what most people working an inbox see: patience shrinks every year.
What do customers expect per channel?
The expectation follows the channel: someone choosing chat expects minutes, someone emailing expects the same day, and someone writing on social media expects a reply within a day.
| Channel | Expectation | What the evidence shows |
|---|---|---|
| Chat | A reply within a few minutes, while the customer has the window open | Gartner expects self-service and live chat to overtake phone and email as the most important channel technologies by 2027 |
| Email and forms | Same working day, and for many customers within a few hours | HubSpot: more than half of managers say customers expect resolution within three hours |
| Social media | Within a day | Sprout Social: 73 percent expect a reply within 24 hours or faster |
| Phone | Someone answering without a queue | Thin evidence in current reports; the expectation in practice is not to wait at all |
The sources for the table are Gartner on channels and the Sprout Social Index 2025 on social media.
Two things to take away. The customer judges you by the norm of the channel, not by your staffing: a chat that answers after twenty minutes feels closed, while an email answered within twenty minutes feels fast. And the same customer switches channel when the first one does not deliver. Someone who gets no reply to the email writes in the chat or on Instagram, and then you have two tickets instead of one.
What is the difference between first reply and resolution?
First response time is the time from the customer getting in touch to a person or an AI giving a substantive reply. Resolution time is the time until the issue is closed and the customer has what they needed. An auto-reply does not count as a first reply, and a first reply that only says "we will get back to you" moves the wait; it does not end it.
Zendesk describes in CX Trends 2026 how the expectation has shifted from a fast reply to an immediate resolution, and that 85 percent of CX leaders say customers leave brands that do not resolve the issue at the first contact. That makes first contact resolution the measure that matters most, and first response time a prerequisite rather than a goal. Both measures, and how to measure them without fooling yourself, are described in the article on customer service metrics.
There is a trap here for anyone measuring. A team that pushes down first response time by sending short acknowledgements gets nice numbers and unhappy customers. Measure first response time on substantive replies, and always measure resolution time next to it.
What happens to loyalty when the customer has to wait?
Waiting costs purchases, not just ratings. According to Zendesk, 86 percent of consumers say speed and a correct resolution strongly influence their buying decisions. On social media the link is direct: according to Sprout Social, 73 percent of users say they will buy from a competitor if the brand does not respond.
Younger customers give up faster. Gartner found that 38 percent of Gen Z and millennial customers are likely to give up on a customer service issue if they cannot resolve it themselves in self-service. They do not call instead. They leave.
For an online retailer the most tangible example is the pre-purchase question. The customer wondering whether the jacket comes in size 40 or whether the parcel will arrive by Friday does not wait for an email tomorrow. If she gets no answer now she buys elsewhere or not at all, and that never shows in the ticket statistics, only in conversion.
What does self-service do to response time?
Self-service is the only channel with zero waiting, but only when the answer exists and is correct. The gap here is wide. According to Gartner, only 14 percent of issues are fully resolved in self-service, and even for issues customers themselves call very simple the share stops at 36 percent. The rest end up in the queue anyway, now with an irritated customer who has already tried.
That means two things for response time. First: every recurring question that gets a good answer in the help centre, chat or form disappears from the queue and shortens it for everyone else. Second: self-service that does not resolve the question lengthens the perceived wait, because the customer counts the time from the first attempt. What matters is therefore not whether you have self-service, but whether the content behind it is complete and reviewed. What that takes is described in the article on knowledge bases, and which questions suit which channel in the article on self-service, live chat and email.
Supportifier is built on that principle: the same reviewed knowledge base answers in the help centre, the chat and the form, and hands over to the team when the answer is missing instead of letting the customer wait for a guess.
What to do
- Measure your actual first response time per channel for the last 30 days. Use the median, not the mean, so a few long-running tickets do not hide what most customers experience.
- Measure resolution time and the share resolved at first contact next to it. If first response time is good but resolution time is poor, you are replying fast with too little.
- Set a target per channel based on the channel's norm: minutes in chat, hours in email, within a day on social media. Publish the target where the customer sees it, so the expectation matches.
- Remove the auto-reply from the statistics. It is not a reply.
- List the ten most common questions in the queue and make sure each has a complete answer in the help centre and in the form. That is the fastest route to a shorter queue.
- Look at when tickets arrive per hour and weekday and compare with when you are staffed. Many long response times are a staffing question, not a skills question. How to calculate the need per hour and channel is described in the article on calculating staffing.
- Follow up every week. Response time moves fast, in both directions.
Common questions
Does an auto-reply count as a first reply?
No. An auto-reply confirms the ticket has arrived, but it does not answer the question, and the customer counts the wait until someone does. Measure first response time from when the ticket is created to the first substantive reply, from a person or an AI. An auto-reply can still be valuable if it states the expected response time and links to an answer in the help centre that may resolve the question directly.
Is it better to reply fast with half an answer or slowly with a full one?
Fast and complete is the goal, but if you have to choose: a fast reply that says what will happen and when, followed by a full reply on time, beats both alternatives. What the customer cannot stand is silence. Avoid, though, fast replies that just push the question back to the customer, such as "can you send the order number?" when the order number is already in the email.
Should we promise a response time on the site?
Yes, if you keep it. A published response time, for example "we reply to email within four hours on weekdays", reduces the number of reminders and helps the customer choose the right channel. Promise slightly less than you can deliver. A broken promise costs more trust than no promise at all, and it shows first on social media.
How fast do we have to reply at weekends?
Customer expectations do not follow your opening hours; according to Zendesk, 74 percent expect availability around the clock. That does not mean you have to staff weekends, but that self-service, forms and chat must be able to resolve the most common questions without you, and that the customer is told when a person will reply. For e-commerce, Sunday evening is often the busiest hour of the week for questions.
Sources
- Contextual Intelligence Becomes the New Standard for Exceptional Customer Experience in 2026 (CX Trends 2026) — Zendesk, 2025
- HubSpot Annual State of Service Report — HubSpot, 2024
- Customer service statistics to know in 2025 — HubSpot, 2025
- Social Media Customer Service Statistics for 2025 (Sprout Social Index) — Sprout Social, 2025
- Gartner Survey Finds Self-Service and Live Chat Will Surpass Traditional Channels as Top Customer Service Technologies By 2027 — Gartner, 2025
- Adapting to the Customer Service Preferences of Gen Z and Millennials — Gartner, 2023
- Gartner Survey Finds Only 14% of Customer Service Issues Are Fully Resolved in Self-Service — Gartner, 2024
Martin Carlsson
Rickard Collander