Klyro
Back to blog
November 11, 2026

Response time per channel: expectation and measurement

It presents the expected response time of Brazilian consumers by channel (chat, WhatsApp, e-mail), explains the difference between first response time and resolution time, advises how to measure average time by channel and by type of
Classic vintage mechanical metronome on a wooden table, perfect for music enthusiasts.
Photo by Pixabay on Pexels

What Brazilian consumers expect from each channel

Expected response time varies by channel. According to data from the customer experience manual, ** 36% of consumers expect an online chat response within 1 minute**, while ** 26% accept up to 5 minutes**. On the other hand, in messaging apps like WhatsApp, ** 19% expect a response within 1 minute** and ** 32% within 5 minutes**. Email offers more tolerance: only ** 9% expect a response within 1 minute**, ** 16% within 5 minutes**, but ** 48% accept up to 24 hours**.

This difference is not arbitrary. Chat on the site is associated with urgency and navigation; the customer has not yet left the page. WhatsApp is tied to convenience and to the ongoing conversation; the customer may be engaged in another activity but expects a return soon. E-mail carries the expectation of a formal process, with a structured response.

Defining SLA by channel without considering this variation creates two problems: either you commit resources disproportionately to the wrong channel, or you frustrate the customer by not meeting the natural expectation of the medium they chose.

First answer versus full resolution

Average response time per channel can mean two different things: the time until the first message from the attendant or the time until the problem is solved. Confusing the two leads to unrealistic goals and metrics that reveal no bottlenecks.

The first response is the moment when the attendant enters the conversation. It's critical to reducing abandonment and frustration, but it doesn't solve the customer's problem. The complete resolution includes the entire process: investigation, consultation of systems, possible transfers between areas, and confirmation that the demand has been met.

Measuring only the first response can hide conversations that drag on for days across sectors. Measuring resolution alone can punish attendants who respond quickly but rely on other areas to complete. The ideal is to monitor the two metrics independently: first response time per channel and total resolution time by type of demand.

How to measure first response time per channel

To measure accurately, record the opening of the conversation at the time the customer sends the first message and the first response at the moment when a human attendant or AI agent responds in a contextualized manner. Automatic confirmation of receipt responses don't count as the first actual response.

Organize the measurement by source: WhatsApp, Instagram, chat, email, and phone should appear separately in the report. Calculate the average, but also look at the median and the 90th percentile. A low average can hide conversations that go unanswered for hours while others are answered in seconds.

If each channel is on a different tool, consolidation becomes manual labor and the number is late. Omnichannel platforms such as [unified inbox] (URL0) were designed to centralize WhatsApp, Instagram, Facebook, Telegram, email, and website conversations, allowing first response time to be measured automatically and comparably across all channels.

Identify causes of delay without opening conversation by conversation

Excessive delay rarely has a single cause. The most common bottlenecks are: a waiting line with no clear distribution, lack of customer information that forces the attendant to return several times, transfers between sectors that reset the context, and the absence of a defined responsible person to follow up to the end.

To identify where time is being wasted, analyze conversations that exceeded the SLA and map at which stage the delay occurred. If the problem is with the input, the screening needs to be reviewed. If it's in the middle, the transfer flow between areas is broken. If it is at the end, there is a lack of structured follow-up.

Tags applied automatically by rules or AI help to categorize demands in real time, allowing you to measure volume and average time per reason of contact without having to open ticket by ticket. When the average time for a specific category skyrockets, you know where to review the process before hiring more people.

Strategies to reduce response time without expanding staff

Reducing response time isn't just a matter of adding attendants. Reorganizing the entry of demands, eliminating rework, and preserving context between transfers usually has more impact than doubling the team.

Start by screening. A well-configured bot or AI agent collects basic information before sending the conversation to the right sector, reducing back and forth. According to CX Trends 2026, ** 63% of consumers say that for a robot to serve well, it needs to connect them to a human when required**, which means that automation works when it anticipates context, not when it tries to solve everything on its own.

Then, unify the history. When the customer needs to repeat information when changing channels or attendants, the total resolution time increases and satisfaction decreases. Platforms that centralize channels and maintain context per customer allow any attendant to continue the conversation where they left off, without starting from scratch.

Finally, define those responsible by type of demand. Conversations that need follow-up - exchange, refund, deployment - must have a clear responsible person and a return deadline. Without this, demand stands still between sectors and the customer charges again, generating more volume.

Difference between SLA by channel and SLA by type of demand

SLA per channel defines how long the customer can wait on each communication medium. SLA by type of demand defines how long each category of problem may take to resolve. The two need to coexist.

A simple question about operating hours should be answered within minutes, regardless of the channel. But an exchange request can take hours or days to complete, even if the first response was quick. Defining only SLA per channel ignores the complexity of demand; defining only SLA by type ignores the expectation associated with the medium.

In practice, you need two follow-up lines: first response time per channel (WhatsApp within 5 minutes, email within 24 hours) and resolution time by category (simple question within 1 hour, exchange within 48 hours). When a specific category starts to run out of deadline, you know that the problem lies in the process of that demand, not in the channel.

The impact of the chat window on WhatsApp

The WhatsApp Business API imposes a 24-hour window for free replies after the customer's last message. Outside this window, the company can only send messages through templates approved by Meta, which are charged per conversation initiated.

This means that responding quickly on WhatsApp isn't just a matter of experience; it's also a matter of operating cost. If the attendant takes more than 24 hours to give the first return, the company must pay to reopen the conversation. And if the customer has to ask again because the answer was not complete, a new window opens and a new charge can be generated.

Reducing the number of messages exchanged per conversation, providing complete answers at the first interaction, and preventing the customer from having to charge follow-up are strategies that impact both the experience and the cost of operation. To understand the billing model in detail, see [service messages on WhatsApp: what is charged] (URL1).

How to set up real-time SLA alerts

Monitoring the SLA after the deadline has passed does not change the result. The ideal is to configure alerts to warn when a conversation is close to the limit, allowing the manager to redistribute demand or prioritize service before the deadline is missed.

Set automatic triggers: when a queued conversation exceeds 80% of the SLA without attribution, notify the coordinator. When an assigned conversation exceeds 90% of the SLA without a response, notify the attendant and supervisor. When a conversation transferred between sectors is left without assuming responsibility for more than 30 minutes, step up to management.

These rules must be configured in the system, not in spreadsheets or manual reminders. Omnichannel service platforms allow you to define alerts by channel, by queue and by type of demand, ensuring that each conversation has a clear responsible person and a deadline followed up in real time.

Qualified deflection strategy to reduce volume

Qualified deflection doesn't mean pushing the customer to a generic FAQ. It means resolving repetitive demands in an automated manner, freeing the team for interactions that require human judgment.

Identify the five most frequently asked questions by channel. Set up contextualized automatic responses, with buttons that take the customer to the right information or to a knowledge base article. When automation doesn't work, transfer with the context preserved: what the customer asked, what the bot answered, and why the transfer was necessary.

Measuring deflection by total volume is misleading. What matters is the qualified deflection: how many conversations were resolved without human intervention and with the satisfied customer. If the return rate within 24 hours is high, automation is deflecting the problem, not solving it.

Average time analysis per person in charge and per queue

Overall average time hides inequalities. One attendant may be solving simple demands in minutes while another is stuck in complex cases that take hours. One row may be scaled and another may be overloaded.

Analyze average response time and average resolution time per attendant and per queue. Compare with the volume of conversations answered in the same period. If an attendant has a short time but solves few cases, they may be prioritizing quantity over quality. If a queue has high time but proportional volume, the problem may lie in the complexity of the demands, not in the capacity of the team.

Real-time dashboards such as those available in the [Analytics] module (URL2) of BRANDallow monitoring of these metrics by channel, by agent, and by period, facilitating decisions based on real data and not on perception.

Cross-channel integration and context preservation

When the customer starts a conversation on Instagram, continues on WhatsApp and closes via email, each step must be connected. If he has to repeat the problem on each channel, the total resolution time triples and the frustration increases.

Omnichannel platforms unify conversations from different sources into a single history per customer, allowing any attendant to see what has already been said, on which channel and with what result. BRANDwas designed to centralize WhatsApp, Instagram, Facebook, Telegram, email, and website conversations on a single platform, ensuring that the context is preserved regardless of where the customer contacts.

This does not eliminate the need to define SLA per channel, but it ensures that the time spent on each stage is visible and that transfers between media do not generate rework.

Frequently Asked Questions

**How to set realistic SLA for each channel without compromising operation? ** Start by mapping customer expectations based on research and market data, such as those presented in this article. Then, measure the team's current capacity: how many conversations each attendant can solve per hour on each channel. Define the SLA at the balance between expectation and capacity, and review it quarterly based on actual compliance data. If the SLA is being met in 95% of the cases, you can tighten. If it's being breached by 30%, you need to review the process before adjusting the goal.

**What is the difference between first response time and resolution time? ** First response time is the interval between the customer's message and the first contextualized human or AI interaction. Resolution time is the total interval until the demand is completed and the customer confirms that the problem has been addressed. Measuring only the first response can hide bottlenecks in the middle of the process; measuring only the resolution can punish agents who depend on other areas. Track both metrics independently.

**How do you know if automation is helping or hindering response time? ** Measure three indicators: deflection rate (how many conversations were resolved without human intervention), return rate (how many customers return to ask within 24 hours), and satisfaction at the end of the automated interaction. If the deflection is high but so is the return, automation is deflecting the unsolved problem. If satisfaction is low, the bot is responding quickly but inadequately. The objective is to solve in an automated and satisfactory manner, not just to reduce volume.

**Why can measuring overall average time hide operational problems? ** Because the overall average mixes simple and complex demands, fast and slow channels, and experienced service providers with beginners. An average of 2 hours can hide conversations resolved in 10 minutes and other stops 3 days ago. Analyze by channel, by type of demand, by queue, and by attendant. Also use the median and 90th percentile to identify extremes that the average hides.

**How to prevent transfers between sectors from increasing total resolution time? ** Define clear transfer criteria, preserve the context of the conversation (history, data collected, stage of the process), and ensure that the receiving sector has visibility of what has already been done. Set up alerts for transferred conversations that go longer than 30 minutes without taking responsibility. Use a platform that allows transfers with full context, preventing the customer from having to repeat information.

**Is it worth investing in omnichannel just to reduce response time? ** Omnichannel isn't just about speed; it's about continuity. When each channel is on a different tool, the history is lost, the customer repeats information, and the total resolution time increases even if the first response is quick. Centralizing channels allows any attendant to take over the conversation from where they left off, reduces rework and improves the experience without necessarily increasing the team. To understand how to unify channels, see [unify Instagram and WhatsApp in one place] (URL3).

References

  • [Customer Experience Manual: How to Set Up Your Operation Manual | Octadesk] (URL4)
  • [Customer service: what it is, what it does and how to organize | Octadesk] (URL5)
  • [How to Safely Automate WhatsApp Calls | Huggy Blog] (URL6)
  • [Fixed number in WhatsApp Business: how to validate and when to migrate | Octadesk] (URL7)
  • [Service URA: how to direct calls by sector] (URL8)
  • [AI agent vs Chatbot: what's the difference and which one to use?] (URL9)
  • [Reclame Award HERE: how it works and what it measures in service | Octadesk] (URL10)

Share

Open Instagram and paste the link in your post/story.