Free Service Level Agreement Template for Maintenance and Field Service (PDF)

The client's facility manager says the P1 response was 40 minutes late. Your technician says he was at the gate in 55 minutes and spent the next 45 waiting for security to find the plant room key. Both are telling the truth, and the contract doesn't say who's right, because nobody wrote down whether the clock stops while you're waiting to be let in.

Most SLA arguments in maintenance work come from gaps like that one. This template is written to close them before they come up. It's built for contractors who send technicians to buildings and equipment, so "response" means a qualified person on site, and the clauses cover the things that cause disputes in practice: when the clock starts, what pauses it, who has to provide access, and what happens when a target is missed.

What's in the template

The agreement itself runs to about eight pages. The details that change most between clients (site lists, assets, the priority matrix) sit in schedules at the back, so you can reuse the main text and only update the schedules for each new contract.

How to fill in each section

The notes follow the template clause by clause. Clauses 1 to 3 are mostly admin, so if you're short on time, skip ahead to priority levels.

Parties, scope and exclusions (clauses 1 and 2)

Most of clause 1 is names and dates. The line people skip says which document wins when the SLA and your main service contract disagree, and it's the one you'll want if the relationship ever turns sour. Say the SLA caps service credits at 10% of the monthly fee, and the main agreement, signed two years earlier, has a penalty clause with no cap at all. Without that line, which one applies is whatever the client's lawyer can argue. The usual split leaves money and legal terms to the main contract and performance to the SLA.

Keep sites and assets out of the body of the agreement. They go in Schedules A and B. When the client buys another building in March, you add a row and nobody has to sign anything again.

List your exclusions in full. In building services contracts they often cover:

• equipment that's beyond economic repair
• damage caused by tenants or other third parties
• work that needs specialist access equipment
• systems maintained by another contractor

If an exclusion isn't written down, expect the client to assume the work is included.

Service hours and contact channels (clause 3)

The hours take one line. The template's example reads "Monday to Friday, 07:00 to 18:00, P1 out of hours on the emergency number", and yours will probably look much the same.

Spend longer on the channels. Facility managers text technicians all the time, usually because they saved the number on the last visit. Say one of those texts lands at ten past six on a Friday and the technician reads it on Monday morning. The client counts from Friday. You count from Monday. Clause 3 heads that off. Requests that come in on the helpdesk line or through the client portal start the clock the moment they arrive. Anything sent another way starts it when you reply.

Priority levels (clause 4 and Schedule C)

Spend more time on this with the client than on anything else in the document. Base each priority on consequences, and give examples the client's own staff will recognize. Schedule C in the template starts from these suggested values:

Treat the numbers in Schedule C as a starting point for the conversation with your client. Nobody has set an industry standard for maintenance response times. The useful test is whether you can staff a target on a wet Friday night in February, when two technicians have called in sick.

If you'd like to compare with a real contract, the University of Lincoln publishes the SLA its estates maintenance service works to. Its emergency tier asks for the contractor to attend within one hour and restore service, or provide alternative facilities, within 24 hours. Its urgent tier allows four hours to attend during working hours.

Clause 4 also needs to say who sets the priority. That's usually your helpdesk, working from the client's description of the fault. It then needs a way to settle arguments about it. Under the template, the technician can change the priority after arriving, as long as they leave a note on the work order. If the client disagrees, they can escalate to the named contact in Schedule E.

Response, make safe and resolution (clause 5)

Three different moments, and an SLA that blurs them will be argued over.

• Response in the template means a technician qualified for the work arrives on site and records the arrival on the work order. A phone call doesn't count, unless the priority matrix says a remote first response is acceptable for that level. If you want phone triage to count for P3 jobs, write it into Schedule C explicitly.
• Make safe means the immediate risk is dealt with: the water is isolated, the area is cordoned off, the trapped passengers are out. It's often all that's possible in the first few hours, especially when a part has to be ordered.
• Resolution is the point where the fault is fixed and the client has signed it off. A temporary repair can count too, if the client agrees to it in writing.

Keeping three separate targets means the monthly report can show what actually happened.

Take a burst pipe. The technician is there in 40 minutes and has the water isolated ten minutes later. Then the replacement valve takes five days to arrive. That's a response target met and a resolution target missed, and the client should see both. A single "SLA met" figure would hide the second one.

How the SLA clock works (clause 6)

This clause prevents more arguments than any other. Here's the wording in the template, with the reasoning behind each part.

• When it starts: "The response clock starts when a request is logged through an approved channel in Clause 3, as shown by the Provider's work order timestamp. Requests sent outside approved channels start the clock when the Provider acknowledges them." Using the work order timestamp gives both sides a single, checkable record.
• What counts as a response: "A request is responded to when a technician qualified for the work arrives at the site and records arrival on the work order." "Qualified" stops anyone sending the nearest person just to stop the clock.
• When it pauses: "The clock pauses while (a) access is refused or unavailable, (b) a quote awaits the Client's approval, or (c) parts are on order and the Client has agreed in writing to the delay. The Provider records each pause and its reason on the work order." This is the clause that settles the plant room key argument from the start of this article. The condition about recording each pause matters: without it, every late job acquires a pause after the fact.
• When it stops: "The resolution clock stops when the Client or its representative signs the completion record, or [48] hours after the Provider submits it without a recorded objection." The second half covers the client contact who's never available to sign.
• Reclassification: "Either party may reclassify a priority within [1] hour of the first response. The target for the new priority applies from the original start time." If a P3 turns out to be a P2, the stricter target applies from when the request first came in. If a P1 turns out to be a P3, you aren't penalized for having treated it as urgent.

Whatever software you use to track SLAs has to handle pauses. The technician should be able to record on the work order when the clock stopped and why. Plenty of tools only count down, and a countdown with no pause will log breaches the contract says never happened. Our guide to field service management metrics looks at the same counting rules from the reporting end.

Planned and statutory maintenance (clause 7 and Schedule D)

Most free SLA templates stop at reactive work. For a maintenance contract that's a strange gap, because planned visits are often most of what the client is paying for. This template gives them their own service level.

Schedule D lists each planned task, how often it's due and how far it can move. Two weeks either side of the due date is usually fine for a quarterly filter change. Statutory inspections are different. A periodic lift examination can't slide at all, and the template marks tasks like that as zero-tolerance. It also asks you to agree a notice period before planned visits, five working days for example, so the client has time to arrange access.

You're measured on PM compliance, meaning the share of planned visits finished inside their window. Our preventive maintenance checklist goes through what each visit should cover. Automated scheduling handles the other half of the problem, which is making sure every visit in Schedule D ends up in someone's calendar.

Provider and client responsibilities (clauses 8 and 9)

It's tempting to write an SLA that's all about the contractor. Clients like it, and it gets signed quickly. Then a quarterly review comes round and some of the missed targets turn out to be jobs where nobody could find the plant room key, or where a quote sat unapproved for two weeks. Your response times depend on the client doing a few things too, and clause 9 writes them down.

On the client's side, the template asks for site access during agreed hours, safe working conditions, one named contact per site with a deputy, approval or rejection of quotes within a set number of working days, and fault reports that include the location and a description. None of this is unusual, and most facility managers will agree to it readily once it's written down.

On your side, clause 8 covers qualified and certified staff, safe working, and evidence of work. The evidence part is worth tightening. The template requires a completion record for every job with what was done, photos where relevant, parts used and the client's sign-off, sent within a set time. Clients who receive that consistently tend to stop disputing invoices, because the proof arrives with them.

Subcontracted work (clause 10)

If you subcontract any trades, such as lifts, specialist fire suppression or out-of-area callouts, clause 10 keeps you accountable for them. Subcontracted jobs follow the same clock, evidence and reporting rules as your own. You can also agree to notify the client of which subcontractors you use for specific trades.

The practical problem is visibility. Subcontracted jobs often disappear into email until the invoice arrives, which is a bad time to discover a P2 was missed three weeks earlier. Putting subcontractors on the same work order system solves this, as long as the system lets you hide your client rates from them.

Service credits (clause 12)

Service credits are a discount on the monthly fee when a target is missed. They're a commercial choice and a legal one, so agree the numbers carefully and have them reviewed. The template uses illustrative values:

• 2% of the monthly fee for each missed P1 target
• 1% for each missed P2 target
• no credits for P3 and P4, which are reported but not credited
• a cap of 10% of the monthly fee in any month
• earn-back: credits for a month are waived if the next two months both meet an agreed overall compliance level

The cap protects the contractor from a single bad week wiping out the month's margin. The earn-back gives both sides a reason to fix the cause instead of arguing about the credit. Clause 12 also has a placeholder for whether credits are the client's only remedy for missed targets, which is a question for your lawyer.

Exclusions, changes and signatures (clauses 13 and 14)

Clause 13 lists situations where targets don't apply: severe weather, utility failures outside the site, obsolete parts with long lead times, and delays caused by the client or other contractors. Clause 14 sets an annual review of the targets, which is worth keeping even when things are going well, since a contract's first-year targets are often guesses.

Worked example: one month under the SLA

The filled-in example uses a fictional contractor, Northline Building Services, maintaining HVAC, electrical and plumbing systems across a 12-site office portfolio for a property manager. Here's how one month plays out against the template's targets. All figures are illustrative.

In March, Northline logged 140 requests:

Two of the jobs are worth a closer look.

• The P1 that looked late: A water leak at site 7 was logged at 22:10. The technician arrived at 23:35, which is 85 minutes against a 60-minute target. But the work order shows a pause from 22:50 to 23:25: the site's out-of-hours security guard couldn't be reached, and the technician recorded each attempt. Under clause 6(a), that's 35 minutes of paused time, so the measured response was 50 minutes. Target met, and the evidence is on the work order if the client asks.
• The two P2s that were missed: Both were on the same Monday morning after a cold snap, when heating faults arrived across four sites within an hour. No pause applies, so they count as breaches. Under clause 12, each costs 1% of the monthly fee. On a fee of €18,000, the service credit is €360, well under the 10% cap. The more useful outcome is the conversation at the quarterly review: should Northline keep an extra technician on call after a cold-weather forecast?

Notice what the table doesn't show: a single blended "SLA compliance" figure. Averaged across all 140 requests, March looks like 96.4%, which hides the fact that both breaches were on urgent jobs. Report each priority separately.

Download the fillable SLA template (PDF)

How to report SLA performance

Schedule F in the template sets out a monthly report. Keep it to two pages if you can. Most facility managers want to see:

• response and resolution compliance for each priority, with the number of jobs behind each percentage
• every breach listed, with the reason
• every pause listed, with the reason (this is what makes the pause clause credible)
• PM compliance, with statutory inspections shown separately
• recurring faults: assets with three or more reactive calls in the quarter
• open jobs older than 30 days and what they're waiting for

The report should be built from work order timestamps, not compiled from memory or a spreadsheet at month-end. If it's put together by hand, it takes days and the client has no way to check it. When it comes straight from the job records, anyone can trace a figure back to the jobs behind it.

NEWSEC, which manages commercial office properties, went from three days of manual SLA report compilation to a single click once the data came straight from completed work orders. Most FSM reporting tools can schedule the report to go to the client automatically each month.

Once a quarter, sit down with the client and go through the trends rather than the individual jobs. That meeting is where targets get adjusted, recurring faults turn into recommendations for repair or replacement, and renewals are quietly won or lost.

Five mistakes that make SLAs hard to enforce

• No definition of when the clock starts: Without it, the client measures from when they noticed the fault and you measure from when you saw the email. Both numbers are defensible and they never match.
• One blended compliance figure: A 95% average can hide a month in which every P1 was late. Report and credit each priority separately.
• Targets nobody checked against staffing: A four-hour P2 response across 40 sites in two cities needs a certain number of technicians on shift. Work that out before you sign, not after the first breach.
• Obligations on one side only: If the client has no duty to provide access or approve quotes, every delay they cause counts against you.
• Uncapped penalties: A credit per missed job with no monthly limit can turn one bad storm into a loss-making month. Cap it, and add an earn-back.

We go further into choosing realistic targets and handling breaches in our guide to service level agreement best practices.

FAQ

What should a service level agreement include?

At minimum: the services, sites and assets covered; service hours; priority levels with response and resolution targets; how the clock starts, pauses and stops; responsibilities on both sides; how performance is reported; and what happens when targets are missed. For maintenance contracts, add planned and statutory maintenance windows. The template above covers all of these in 14 clauses.

What's the difference between an SLA and a service contract?

The service contract usually covers price, term, liability and termination. The SLA sets measurable performance levels and how they're reported. They're often separate documents, with the SLA attached as a schedule to the contract, so state which one takes precedence if they conflict.

What's a typical response time in a maintenance SLA?

There's no industry standard, and published examples vary. The University of Lincoln's public SLA asks its contractor to attend emergencies within one hour and urgent faults within four hours during working hours. Base yours on what you can staff across the client's sites, and agree it priority by priority.

What's the difference between response time and resolution time?

Response time runs from the request being logged to a qualified technician arriving on site. Resolution time runs until the fault is fixed and signed off. You can meet one and miss the other, so measure them separately.

Are SLA service credits legally enforceable?

It depends on your jurisdiction and how the clause is worded. In some legal systems, a credit that looks like a penalty rather than a genuine pre-estimate of loss can be challenged. Have a lawyer review clause 12 before you use it.

After it's signed

Writing the SLA is the easy part. The hard part is proving every month that you met it, with timestamps the client trusts, pauses recorded as they happened, and a report that doesn't take three days to put together.

Serfy runs SLA countdowns and escalations on work orders, the technician's app works with no signal, and SLA compliance reports come straight from the job records. See how SLA reporting works in Serfy or start a free trial.