Customer service SLA: what it is and how to set realistic deadlines
SLA sounds like something out of a multinational contract, but the idea fits in one sentence: it is your support team’s deadline promise. “We answer questions within 2 business hours.” “A problem that blocks a sale gets fixed the same day.” That is an SLA (Service Level Agreement)—and small businesses gain a great deal by adopting one, because it trades “I’ll reply when I can” for a commitment both the team and the customer know.
Why an SLA changes support
- For the customer, an SLA turns waiting into predictability. Waiting 4 hours knowing the answer comes within 4 hours is one experience; waiting 4 hours in a vacuum is an entirely different one.
- For the team, an SLA turns the queue into a clear work order. Without it, each agent decides alone what to do first—and the criterion becomes “whoever complained loudest”.
- For the manager, an SLA creates the yardstick. “Are we serving customers well?” stops being an opinion and becomes a number: the percentage of tickets resolved within the deadline.
How to set realistic SLAs in 4 steps
1. Split by category, do not promise a single deadline
The most common mistake is promising the same deadline for everything—which makes it either too loose for urgencies or too tight for everything else. Start with 3 or 4 ticket categories (question, order/quote, problem, complaint) and give each one a deadline.
2. Add priority
Within each category, not everything is equal: a problem that stops the customer from buying cannot wait like a cosmetic tweak. A simple priority scheme (low, medium, high, critical) combined with category already covers the vast majority of cases.
The priority criterion has to be objective, or it turns into an argument. A yardstick that works: how much of the customer’s business is stopped? Cannot sell is critical; annoyed is medium; would prefer it differently is low. That moves priority out of the realm of likeability and into the realm of impact.
3. Look at your history before promising
A good SLA is one you meet. If your team currently answers questions in half a day, promising 1 hour manufactures frustration. Measure the real response time over the past few weeks and promise something slightly better than the current average—you tighten it later, once the process matures.
4. Separate first response from resolution
These are two distinct deadlines, and both are worth tracking: first response (“we got it, we’re on it”) and resolution. A fast first response, even without a solution, already eliminates the customer’s worst feeling: having spoken into the void.
A starting table
| Category | First response | Resolution |
|---|---|---|
| Question | 2 business hours | 4 business hours |
| Order / quote | 4 business hours | 1 business day |
| Problem (high priority) | 1 business hour | same day |
| Complaint | 2 business hours | 2 business days |
Adjust the numbers to your reality—the value lies in having the table, communicating it, and measuring it.
The detail that prevents half the arguments: defining “business hour”
It sounds pedantic, but it is the number one source of SLA misunderstandings. If you operate 9am to 6pm and a ticket arrives at 5:30pm with a 4-business-hour deadline, it is due at 12:30pm the next day—not 9:30pm tonight.
Write down three things before promising any deadline:
- Your operating hours (and whether Saturday counts).
- What happens to requests arriving outside those hours (the clock starts at the next opening).
- Holidays, which suspend the count.
Without this, half the disputes about “you missed the deadline” are really disputes about which clock applied.
What to do when the deadline is missed
It will be missed. No SLA is met 100% of the time, and promising that is the fastest way to lose the credibility of your own yardstick. What separates a mature operation from a mess is what happens after the breach:
- Warn before it expires, not after. “Your ticket will miss the deadline, the new estimate is X” preserves trust; silence destroys it. This is the most important rule on the list.
- Escalate priority automatically. A breached ticket cannot stay at the back of the queue—otherwise a delay becomes abandonment.
- Move it to somebody else. If it stalled with one person, pushing the same person rarely changes the outcome.
- Log the reason. Breaches caused by volume, by waiting on a third party, and by lack of knowledge call for completely different fixes.
Communicating the SLA is half the benefit
An SLA that exists only internally organizes the team but throws away the bigger gain: the customer’s expectation. And a miscalibrated expectation is the source of most support dissatisfaction—including on tickets where the deadline was actually met.
Three cheap places to make it explicit:
- In the automatic acknowledgment. “We received your message. Questions are answered within 2 business hours.”
- On your site’s contact page, alongside your operating hours.
- In the email signature of whoever handles support.
The effect is immediate and slightly counterintuitive: promising a longer deadline and meeting it produces more satisfaction than promising a short one and missing it. The customer does not compare your deadline against your competitor’s—they compare it against what you promised.
How to measure: three numbers and one trap
- % within deadline (tickets resolved inside SLA ÷ total). This is the headline indicator.
- Average first response time. What the customer feels most.
- Average resolution time. What reveals process bottlenecks.
The trap is looking only at the average. A 3-hour average can hide 90% of tickets answered in 20 minutes and 10% abandoned for two days—and it is precisely those 10% that generate the complaint, the bad review, and the lost customer. Also look at the worst case of the week: it tells the story the average conceals.
Realistic targets—and one impossible target
Aiming for 100% within deadline from the start is the shortest path to a team gaming the number (closing tickets that were not resolved, reopening them as new, classifying everything as low priority). A target of 90% to 95% is demanding and honest, and leaves room for the unexpected that always shows up.
And the uncomfortable question is worth asking: if you have been hitting 100% comfortably for months, the SLA is too loose to mean anything.
An SLA without tracking is decoration
Setting the deadline is half the work; the other half is seeing, in real time, what is close to breaching. That is where manual control collapses: nobody can watch dozens of conversations with a stopwatch in hand. A ticketing system does it for you—each ticket carries its deadline, and the board shows what is green, yellow, and red.
If your support still lives entirely inside a messaging app, the step before defining an SLA is getting control out of there—a deadline over an invisible queue is not a deadline, it is an intention.
In Mova Helpy, the SLA is configurable by category and priority exactly as described above, with a real-time tracking board and team performance reports. Plans start at R$ 89.90 per month, with a 7-day free trial and no credit card. You define the promise; Helpy watches the clock.
Frequently Asked Questions
What does SLA stand for in customer service?
SLA (Service Level Agreement) is the target response and resolution timeframe agreed upon for customer inquiries.
What is the difference between first response time and resolution time?
First response time measures how fast an agent acknowledges the request; resolution time measures how long it takes to solve the issue completely.
What is a good SLA for a small business?
A common starting point: high priority resolved within 4 business hours, medium within 1 to 2 business days, low within 3 to 5 business days. Start with deadlines your team already meets today and tighten gradually.
What should happen when an SLA deadline is breached?
The ticket should escalate in priority and move to someone else, and the customer needs an explanation. Silence after a missed deadline erodes trust more than the delay itself.