The practical guide to XLAs.

What to put behind the target so employee experience improves: continuous feedback, happiness and lost time, and a review rhythm that leads to action.

Download the PDF

Chapter 1

The target that nobody acts on.

Your SLA dashboard is green and your XLA has a number against it. Employees can still lose hours to incidents, repeated hand-offs, and workarounds nobody sees.

A commitment needs a measurement system.

An experience level agreement commits IT and its providers to human outcomes. It defines which metrics matter and what level counts as good. On its own, it cannot explain where employees lose time, which groups are worst affected, or what to fix first.

Across the HappySignals benchmark, employees lose an average of 3 hours 18 minutes per IT incident. Closing that gap takes trustworthy feedback, service context, and a way to act on the findings.

Where XLAs came from.

Tickets closed on time and dashboards stayed green while employees remained frustrated. This watermelon effect exposed the limits of measuring process compliance alone.

XLAs add a written commitment to the employee outcome. Keep your SLAs for operational performance, and give your XLA the continuous measurement it needs to guide improvement.

SLAs measure the output. XLAs measure the outcome.

Run them together. One tells you the process ran, the other tells you whether it helped.

Question	SLA	XLA
What it measures	The output of IT	The outcome for the employee
The unit	Processes and ticket targets	Added value and productivity
The picture it paints	High-level objectives that are easy to meet	What employees experience and need
Project success	On time and on budget	Did the change make people more productive
The lever	Sanctions	Rewards
The target	Stays the same	Rises as experience improves

Keep your SLAs. Add XLAs to see what they cannot.

“Agreed” does not mean improved.

Changing the agreement's name leaves the same problems in place if the governance, review cadence, and people involved stay unchanged. Before setting a level, name the specific experience problem you want to solve.

Three habits to leave behind.

Designed by IT, for ITInvolve employees in defining what good looks like. Their experience should shape the agreement.
Written as a penaltyReward measurable improvement so providers have a reason to improve the experience.
Applied as a templateSet targets for the services and employee groups you manage. A global average can hide the people struggling most.

A CIO working without Experience Data would be like a Sales Manager working without CRM.

Christopher Woods, CIO, Campari Group

Where an XLA starts

The average meets the target. The frontline does not.

An XLA is a target on experience, not on process: happiness and lost time, by group. The overall score here is +49, so an XLA written on the average would be met. Click a role, a provider, or a country and the groups behind the average appear. The red ones are the groups an XLA should be written for, with their own target and their own review.

Selection +49 3,388 responses

Last 60 daysvs.Baseline +49

Channel

Service Desk Provider

Country

Role

Click any group to drill down. Switch to Lost Time: a good XLA covers both metrics. Illustrative data. Every value is coloured against the selection baseline: green at or above, red below.

A rising score needs a consistent measure.

A score can improve because the survey changed, the harder questions disappeared, or feedback was collected at a different moment. Keep the measure consistent so you can distinguish a better experience from an easier target.

Measure happiness and lost time together.

Happiness tells you how people feel about an interaction. Lost time shows what it cost them. A service can have a healthy satisfaction score and still consume hours of an employee's working day.

Review the level as performance improves.

Expectations rise and services change. Review targets at least quarterly, and raise them when the evidence supports it. The purpose is to learn what helps employees and improve from your baseline.

Chapter 2

The people your XLA never hears from.

Your target is only as representative as the feedback behind it. Typical IT satisfaction surveys receive responses from 5 to 10% of employees asked; HappySignals customer response rates average 15 to 40%.

Look beyond the overall score.

Check whether patterns remain visible when you break feedback down by service, location, team, and role. More responses matter because they help you identify the groups whose experience needs attention.

In the benchmark, the 13% of tickets that bounce between teams cause 80% of lost productivity. Employees report around 2 hours 7 minutes lost for a ticket resolved on first contact, compared with 10 hours 35 minutes when it is passed along five times.

Point your XLA at the segments and interactions where that lost time is concentrated.

Build the practice behind the target.

Capture feedback, share the evidence, identify the experience problems, and measure what changes. Keep the same questions so the results remain comparable.

Stage 1
Measure experiences.
Capture short, contextual feedback at each IT interaction, starting with two or three high-volume services.
Stage 2
Share the data.
Give service owners, employees, and providers the same view of happiness and lost time.
Stage 3
Identify improvements.
Connect feedback to service context and employee segments to see where people lose time and why.
Stage 4
Improve and measure again.
Act on the findings and compare the outcome with your baseline, using a consistent measure.

Use the baseline to set your first target, then return to the data as services and employee expectations change.

A target needs service context.

A missed level tells you that something needs attention. To decide what to change, connect the feedback to the ticket, service, provider, waiting time, and number of hand-offs behind it.

That context helps you answer which service caused the frustration, where the delay happened, and why the employee could not get back to work.

Give clients and providers the same facts.

The benchmark shows outsourced first lines resolving 60% of incidents on first contact, compared with 44% for internal desks. Yet employees supported by outsourced desks report 45% more lost time per incident and 135% more per request.

Independent measurement gives both sides a shared view. Employees answer the same questions against the same benchmark, so service reviews can focus on improving the experience.

Rather than technology-driven, we needed to be experience-led.

Shaun Brown, Director of Global Service Delivery, GEA Group

Chapter 3

What a good XLA looks like.

Start with measurement, agree the outcome, and set a level grounded in your own baseline. Then make the target visible where service owners and providers work.

Measure before you commit.

Capture feedback at each interaction, starting with two or three high-volume service areas. The guide recommends two months of continuous data for a first baseline.

Agree what you are optimizing for.

Define happiness and lost time for the segments you manage. Make the employee outcome concrete before setting the level, because that definition is what your provider will work toward.

Set the level against your own baseline.

Use a benchmark to calibrate ambition and your own baseline to set the number. One shows how other organizations perform; the other shows where you are starting.

Make the target visible and reward improvement.

Put the target in the dashboards service owners and vendors use daily. Pair it with rewards for measurable improvement, alongside any penalties for missed commitments.

Review it on a rhythm.

Review the data monthly, the targets quarterly, and the agreement's structure annually. HappySignals has XLA management built in, so targets, experience data, and review can sit together.

Illustrative target: one service, measured against its baseline.

Experience Indicators out-of-the-box

A score tells you how people feel. Experience Indicators tell you why.

After rating, employees pick what made the experience good or bad from a short list. HappySignals has these Experience Indicators ready for most areas IT works in: services, laptops, applications, and AI tools. Set each one against the benchmark and you see what to fix first, and what to keep.

Experience Indicators

Based on 882 responses with Experience Indicators

Show benchmarks

Negative 8%70 responses with Experience Indicators

Service was slow59%52%

I wasn't informed about progress49%40%

My ticket was not solved47%56%

It was not clear how to get help46%40%

I had to explain my case several times40%49%

Service personnel's attitude11%8%

Service personnel's skills10%11%

Instructions were hard to understand9%11%

Neutral 9%82 responses with Experience Indicators

Service was slow27%29%

I had to explain my case several times23%25%

It was not clear how to get help15%12%

I wasn't informed about progress15%8%

Instructions were hard to understand11%9%

My ticket was not solved6%8%

Service personnel's skills2%5%

Service personnel's attitude1%3%

Positive 83%730 responses with Experience Indicators

Service was fast69%71%

My ticket was successfully solved66%67%

Service personnel's attitude52%54%

I only had to explain my case once45%42%

Service personnel's skills45%49%

I was informed about progress38%35%

Instructions were easy to understand35%37%

It was clear how to get help35%32%

Service desk incidents, demo data. Coloured against the benchmark: red where you do worse, green where you do better.

What the data says.

3h 18m

average employee time lost per IT incident

2026 practical guide to XLAs, HappySignals benchmark

13%

of tickets, those passed between teams, account for 80% of lost productivity

2026 practical guide to XLAs, HappySignals benchmark

57%

less lost time per IT incident at Ahlstrom

2026 practical guide to XLAs, Ahlstrom customer case

From an agreement to a habit.

Ahlstrom and the Scottish Government show what changes when experience data becomes part of how IT and its providers work.

Ahlstrom, global fiber materials company, 45 locations

57%less lost time per IT incident

Targets sharpened because the data came first.

Focusing on experience led to +150% in happiness and 57% in productivity.Dr. Ozan Acar, Head of CIO Office, Ahlstrom

Ahlstrom cut lost time per IT incident by 57%, from over seven hours to just over three, across 45 global locations with no physical IT presence in any of them.

Happiness targets are written into every outsourcing contract. IT and its providers review the same data monthly, building on continuous measurement since 2018.

Monthly reviews with providers
Happiness targets in outsourcing contracts
Continuous measurement since 2018

The Scottish Government

£1Msaved in 2024, followed by £600K in 2025

Experience-led changes made savings visible.

The Scottish Government delivered £1M in savings in 2024 and a further £600K in 2025 through experience-led service and process changes.

Experience data revealed recurring rework when tickets were closed without confirming that the issue had been resolved. The improvement came from understanding what happened to employees.

Shared evidence for service improvements
Rework identified from employee feedback

After you book

From this guide to your first XLA.

A conversation first. Then your tools connected and employees answering.

Day 1
A 45-minute call
You talk to a specialist about your services, your data and the questions you need answered.
Next
Connect ServiceNow, Nexthink or BMC Helix
Pre-built integrations, so measurement starts without a data project.
Then
Your first results
Employees answer short, research-backed surveys. You see happiness and lost time, and the reasons behind them.

Why HappySignals

Why not build it yourself?

Most teams start with what they have: a CSAT survey in ServiceNow, Qualtrics, or Microsoft Forms. Building your own takes months, and you still have no benchmark to compare against. Here is what you get from day one instead.

Happiness+44▲ 11 vs baseline

Lost time / month2h 30per person

The whole story behind the experience

Goes beyond a satisfaction score to explain the context behind what people experience with IT. By combining experience, employee, technical, and operational data, you can identify where people are unhappy and the exact reasons why.

You+46

Benchmark+34

Top quartile+58

Validated surveys and benchmarks

Instead of every organization inventing its own CSAT questions and targets, we provide a proven, standardized survey approach with external benchmarking. Results are comparable over time and across organizations, and setup effort drops dramatically.

Consistently high response rates

Designed for people to express how they feel, which produces consistently higher response rates than typical CSAT surveys. Results are more representative, more trustworthy, and usable for analytics.

Happiness by group

ServicesLaptopsAI tools Office+82+71+45 Remote+74+58+39 Frontline+61+36+12 Field sales+55−30+18

Lowest: field sales, laptops−30

Experience by user segment

Understand experience by meaningful employee segments rather than a single average. Spot hidden pockets of poor experience, compare groups fairly, and target improvements precisely.

Experience Intelligence

Which employee groups are most frustrated?

Your AI experience advisor

Just ask. Our AI draws on over a decade of experience insights from millions of responses, surfacing themes, generating reports, and answering hard questions in seconds.

Lost per IT incident3h 18minbenchmark average

Cost per incident€126European labour costs

Measure where productivity is lost

Lost time translates experience into business impact. That makes it easier to prioritize work, justify investment, and speak the language of leadership.

ServiceNowBMC HelixNexthinkMicrosoft TeamsISO 27001:2022

Time to go liveDays, not months

Ready to deploy

Designed to go live fast, without turning experience measurement into a long survey design, integration, and reporting project. Proven methodology, ready-made surveys, benchmarks, certified integrations, and immediate value from day one.

Take it with you.

Get the practical guide as a PDF.

Read the 2026 guide on turning an XLA commitment into a way of working. Share the 11-page PDF with your service owners and providers.

Why an agreement alone does not improve experience
Happiness, lost time, representative feedback, and service context
A practical order for setting and reviewing targets
Ahlstrom and Scottish Government examples
A six-point checklist to get started