Skip to content
Chatbotscape
Editorial flat-vector illustration for How to Set Live Chat SLA Targets
Published 8 min read

How to Set Live Chat SLA Targets

Quick answer: Start with who the promise is for and what you can already deliver, not with a number from another company. Set a first reply target for chat, a next reply target for the rest of the conversation, and a resolution target only where issues run over hours or days. Add a faster tier for urgent or high-value conversations, decide whether the clock follows business hours, and check the target against your staffing at the busiest hour. Then tell customers the promise, let the help desk flag conversations close to breach, and report the share that met the target every week.

Step 1: Decide who the promise is for

An SLA is either a promise to customers or a target for the team, and the two need different care.

TypeWho sees itWhat happens on a missGood starting point
Internal targetYour team onlyA breach shows in reportsMost small businesses
Published promiseVisitors on your site or chat widgetCustomers notice and may complainOnce you hit the internal target reliably
Contract SLAA named customer, in writingWhatever the contract says, sometimes creditsB2B plans, set with whoever signs contracts

Start internal. Run it for a few weeks before you publish anything or write it into a contract. Intercom's own list of uses covers all three: internal support goals, response times agreed with VIP customers, and reply targets for high-value sales leads.

Step 2: Pick the targets for each channel

Choose the smallest set of timers that describes what customers care about.

  • First reply. The wait until a person first answers. On live chat this is the target customers feel most, because they are still on the page. Measure it from the handoff when a bot answers first; see first response time.
  • Next reply. The wait for each later answer in the same conversation. Without it, an agent can send a quick "hi, looking into it" and then go silent. Zendesk and Intercom both offer it; HubSpot's inbox article covers only first reply and close.
  • Resolution. Time until the issue is closed. Useful for email and tickets that run over days, less useful for chats that end in minutes. Zendesk recommends choosing one resolution metric per policy.

Before choosing numbers, pull your current waits for two to four weeks, split by hour of day. Your first target should be one your team already meets on most days at most hours. A target you miss every lunchtime trains everyone to ignore the red badges.

Step 3: Add priority tiers

One target for every conversation is simple but treats a payment failure the same as a question about opening hours. Two or three tiers are enough for most small teams:

TierExample conversationsHow it gets set
UrgentPayment failed, order not arrived, account lockedA rule on keywords, form choice or bot intent
StandardEverything else during staffed hoursDefault
LowFeedback, feature ideasAgent or rule

How each tool picks the tier differs. Zendesk sets targets by the system Priority field, and a ticket with a blank priority gets no SLA at all, so make sure a trigger sets priority when the ticket is created. Policies apply in list order, so put the strictest at the top. Intercom applies an SLA with a workflow step, and only one SLA can be active on a conversation; a later workflow that applies an SLA replaces the earlier one, so check overlapping workflows. HubSpot applies SLAs to all tickets or by ticket priority, and priority-based SLAs need the Priority property filled in. A pre-chat form or the bot's intent is often the simplest way to set the tier. See our Intercom review and Zendesk.

Step 4: Choose the hours and what pauses the clock

Hours. A business-hours SLA only counts the time you are open. Zendesk lets you choose calendar or business hours for each priority; HubSpot offers "Operation hours" or "24/7" and notes that holidays are not yet accounted for; Intercom spreads its targets across your office hours, so a message received at night can be due the next working day. If you choose business hours, say so to customers and let an after-hours chatbot tell night visitors when the team will reply.

A worked example of how business hours move the due time: with a 4 business-hour email target and a 9:00 to 17:00 week, a message at 16:50 on Friday uses 10 minutes that day and the remaining 3 hours 50 minutes on Monday, so it is due at 12:50 on Monday.

Pauses. Decide whether the clock stops while you wait on the customer. Intercom can pause timers while a conversation is snoozed or waiting on the customer, and if an SLA is missed while snoozed, it reopens the conversation. Several Zendesk metrics pause while a ticket is Pending. HubSpot does not exclude tickets in "Waiting on contact" from SLA calculations, so a resolution timer keeps running there. Two notes on bots: Zendesk says reply-time SLAs for live chat must be turned on, otherwise only public replies count, and autoreplies fulfill reply metrics only with its Copilot add-on.

Step 5: Check the target against your staffing

A target is a staffing decision. Whether two agents can answer 80 percent of chats within 2 minutes at your Monday peak depends on how many chats arrive, how long each takes (average handle time) and how many people are online. Enter those numbers in our response time SLA calculator; it uses Erlang C to show the share of chats answered in time and the smallest team that meets your target. The chatbot handoff design guide explains why small teams need visible slack to keep a fast promise. To plan the whole week hour by hour, see forecasting live chat staffing.

If the numbers do not work, you have four levers: more people at the peak hours, a slower target, fewer chats reaching people because the bot answers more, or shorter conversations because the bot collects details first. Queue settings matter too; see managing a live chat queue.

Step 6: Publish it, alert on it and review breaches

  • Tell customers. In the chat widget or the waiting message, state the target in plain words, such as "We usually reply within a few minutes, 9am to 5pm Monday to Friday." Publish only what you meet at your busiest hour.
  • Let the inbox warn agents. Intercom colors badges orange with under 5 minutes left and red when overdue; HubSpot shows due soon and overdue states with limits you set. Teach agents to sort by time to breach at busy times.
  • Report the share met, weekly. Report the percentage of conversations that met each target, not only the average wait, because an average hides the long waits customers remember. Intercom's reports split SLAs into hit, missed and "fixed", meaning the reply came after the deadline.
  • Read the misses. Group missed conversations by hour, tier and topic. Chat tags make the topic split easy. Most patterns point to one cause: a staffing gap at a fixed hour, a topic that needs a specialist, or a tier rule that sends too much to urgent.

Check the arithmetic

An invented example, not a measurement. A shop with two agents sets an internal first reply target of 2 minutes for chats during staffed hours. In the first week, 400 chats reached a person and 340 got a reply within 2 minutes: 340 ÷ 400 = 85%. Of the 60 misses, 42 arrived between 12:00 and 13:00: 42 ÷ 60 = 70%. The shop moves one agent's lunch break and sets the bot to answer delivery-date questions before the handoff. The next week, 380 chats reached a person and 361 met the target: 361 ÷ 380 = 95%. Two weeks do not prove cause, so keep reporting weekly before publishing the promise. These numbers are made up to show the method.

Common mistakes

  • Copying a number from another company instead of starting from your own waits.
  • Counting the bot's instant reply as meeting a human first reply target.
  • A blank priority field in Zendesk, so tickets silently get no SLA.
  • Overlapping workflows in Intercom replacing each other's SLA.
  • Calendar-hour targets with a team that only works weekdays.
  • Reporting only the average and missing the lunchtime spike.

FAQ

What is a good live chat response time SLA for a small business? No source read for this guide gives an industry standard. Start from what your team already achieves at most hours, set a target you meet on most days, and tighten it once you hit it reliably.

Should I set a first reply or a resolution SLA for live chat? First reply and next reply first, because chat customers are waiting on the page. Add resolution targets for email and tickets that run over days.

How many SLA priority levels do I need? Two or three are enough for most small teams: urgent, standard and possibly low.

Should the SLA clock stop outside business hours? For most small teams, yes, as long as customers are told your hours and night visitors are told when to expect a reply.

How do I report on SLA performance? Report the share of conversations that met each target every week, and read the misses grouped by hour, tier and topic.

Do chatbot replies count toward the SLA? Usually not as a human reply. Zendesk counts autoreplies only with its Copilot add-on, so most SLAs measure the wait after handoff.

Sources

  • Zendesk, Defining SLA policies — support.zendesk.com/hc/en-us/articles/4408829459866, read 15 November 2026. Source of the metric definitions, the one-resolution-metric advice, priority-based targets and the blank-priority rule, policy order, calendar versus business hours, Pending pauses and the live chat and autoreply notes.
  • Intercom, Set SLAs for conversations and tickets — intercom.com/help/en/articles/6546152, read 15 November 2026. Source of the uses of SLAs, target types, office hours handling, pause options, reopening on a missed snoozed SLA, badge colors, hit/missed/fixed reporting and the one-active-SLA rule.
  • HubSpot, Set SLAs in the inbox — knowledge.hubspot.com/inbox/set-slas-in-the-inbox, read 15 November 2026. Source of time to first reply and time to close, all-tickets or priority-based SLAs, operation hours versus 24/7, the holiday note, the "Waiting on contact" note and due soon/overdue states.
  • Chatbotscape evaluation methodology. /methodology (continuously updated).