
A quick response is not necessarily a quick fix. When reviewing IT support, ask what starts the clock, what counts as a response and what happens after somebody takes ownership. A promise to acknowledge your request tells you less than a commitment to begin investigating it.
For a small business, the useful question is what staff can do while a fault remains open. Ask your provider to walk through a realistic interruption and explain the response, temporary workaround, updates and escalation. Write down the answers before agreeing the service, rather than interpreting a headline target during an outage.
Separate the promises you are buying
Ask for separate definitions of acknowledgement, first response, restored service and final resolution. Atlassian’s service-management documentation distinguishes first-response and resolution measurements; that is a useful illustration of why one timer cannot explain the whole service.[1] Your own agreement should state which events it measures, without relying on another supplier’s terminology.
For example, have the provider explain whether an automated ticket receipt counts towards its target. Ask what evidence shows that an engineer has begun investigating. If a workaround lets your team continue, request confirmation of whether the ticket stays open until the underlying fault is fixed. Keep each commitment beside its definition.
Check when the clock actually runs
Ask for the support calendar in writing, including UK bank holidays and any arrangements for staff working elsewhere. Atlassian documents configurable conditions and calendars that determine when SLA measurements start, pause or stop.[3] Do not assume that an elapsed hour and an hour counted under your contract mean the same thing.
Use a practical example: your team reports a fault shortly before the agreed support day ends. Ask when the response becomes due and what happens overnight. If you need help outside those hours, confirm the reporting route, covered incidents and charging arrangements. Monitoring outside office hours should prompt a separate question about who will act.
Agree urgency using business impact
Give the provider examples drawn from your working day. Compare one person losing printer access with everybody losing an essential application, then explain any deadlines that change the impact. Ask the provider to assign a priority to each example and show which response commitment would apply.
Record who can raise the priority if the situation worsens, and how staff should report that change. Avoid making every request an emergency. Instead, agree what information helps the helpdesk judge the problem: affected people, unavailable systems, work that has stopped and any safe alternative. Never include passwords in the fault report.
Ask who owns a delayed ticket
A request waiting for your reply may stop accumulating measured support time in some systems; Atlassian gives waiting for customer information as an example of a paused SLA timer.[1] Ask your provider which waiting states pause its own clock, who authorises them and how you can see the reason.
Check the arrangements for faults involving broadband or specialist software suppliers. Request one named coordinating role, a next-update time and a route for chasing missed updates. Your provider may need another company to fix the underlying issue, but you should still agree who communicates progress and confirms whether staff can resume work.
Test the agreement with one incident
Before signing, run a desk exercise using an application your business depends on. Ask your provider to describe the route from the initial report through investigation, a possible workaround and escalation. Have your office manager follow the same instructions. Look for missing contact details or decisions that depend on somebody who might be unavailable.
Then request a sample service report with customer details removed. Ask how it separates first response from resolution, explains paused time and shows requests that remain unresolved. A percentage without definitions will not answer those questions. Use the sample to agree what you will discuss at service reviews, including recurring faults and missed commitments.
Put the answers where staff can find them
Keep a short support instruction alongside the agreement. It should tell staff how to report an ordinary request, how to report a serious interruption and who to contact if progress stalls. Store an accessible copy outside any system whose failure would prevent you reading it, and check the details when contacts change.
If you are comparing proposals, use our quote comparison guide for the wider scope and costs. For response arrangements, bring two real examples of interruptions your team cannot comfortably absorb. Discuss them with BMT’s managed IT support team and ask for the proposed handling in writing. That gives you something concrete to review before you commit.
