← The journalSUPPORT / FIELD GUIDE

Support tickets with SLAs your customers can feel

Define response and resolution targets by priority, run the clock on real working hours, pause it honestly, and review breaches so the promise means something.

Montreal streets with directional camera-motion blur.SUPPORT / SUNDIAL
Photography: Louis Le Pessot / Unsplash ↗ · Sundial motion treatment
THE PRACTICAL TAKEAWAY

An SLA is a promise with a clock. Set response and resolution targets per priority, count only working hours, pause only when you are genuinely waiting on the customer, and review every breach.

An SLA is a promise with a clock

Many companies have service level agreements in their contracts and nowhere else. The contract promises a four-hour response for critical issues; the support team works from a shared inbox; nobody knows how long any customer has been waiting. Customers notice long before management does.

A useful SLA has two numbers per priority. Response time: how long until a person acknowledges the issue and starts on it. Resolution time: how long until it is fixed or closed with an agreed outcome. Both run on a clock the customer would recognize, and both are visible to the people doing the work.

Set it up in the support module

In Sundial's Support module, customer issues arrive as tickets, by email or created by your team. A service level agreement sets the response and resolution times for each priority, and which issues it applies to: a default agreement for everyone, and specific agreements for customers or contract types that pay for faster service.

Each issue then carries its own due times, calculated when it is created. The team sees which tickets are close to breaching and which already have. The customer record shows open issues next to orders and invoices, so the account manager is never surprised by a frustrated call.

Count real working hours

A four-hour response promise made at 6 pm on a Friday does not mean 10 pm on Friday unless you staff for it. Each Sundial SLA has working hours per day of the week and a holiday list, and the clock only runs inside them. Set these to what you actually staff, not what sounds good in a proposal.

If you sell round-the-clock support for some customers, give them a separate agreement with all-day hours and make sure someone is genuinely on call. A promise the clock enforces but nobody is staffed to meet just measures failure precisely.

Pause the clock honestly

Some time on a ticket is not your time. If you have asked the customer for a log file and they take two days to send it, counting those two days against your resolution target is unfair, and teams that are measured unfairly start gaming the measure.

Sundial lets an SLA pause on particular statuses, such as when the ticket is waiting on the customer. The clock stops while the status holds and resumes when they reply. The discipline is to use the pause only for genuine waits. A ticket parked as "waiting on customer" to stop the clock while the team is busy is a breach hidden from the report, and the customer will remember it even if the report does not.

Warranty, visits and the field

For companies that sell equipment, support extends beyond tickets. Sundial records warranty claims against the serial number sold, so you can see whether the unit is under warranty or covered by an annual maintenance contract before agreeing to free work. Serial numbers come from the same Inventory records used at dispatch, so there is no separate warranty spreadsheet to keep in step.

Field work is recorded as maintenance visits: who went, when, what was done and what parts were used. Contracted periodic visits are planned with a maintenance schedule per customer, which generates the visit dates in advance.

Review breaches, not averages

Average response time is a comfortable number that hides the customers who waited far too long. Review the breaches instead: every ticket that missed its response or resolution target in the past week, with the reason. Patterns appear quickly. One product generates most of the critical tickets. One shift is short-staffed. One kind of question takes days because only one person knows the answer.

Fix the cause, not the ticket: write the knowledge down, adjust staffing, or send the product issue to engineering with the tickets attached. Sundial is designed so that an agent can read a new ticket, suggest its priority and a draft reply from similar resolved issues, and leave the decision and the send to a person. The SLA clock keeps running on the human promise either way.