The practical guide to XLAs.

Four chapters on measuring what employees experience, setting targets that mean something, and making IT human-centric.

Download the PDF

Chapter 1

What are experience level agreements?

XLAs mean slightly different things to different practitioners. The common thread: they shift the question from “did we hit the SLA?” to “did we actually help?”

Different definitions of XLAs

Here is how some of the most respected voices in IT service management define them:

Giarte, a research and consultancy company, defines XLAs as less like traditional agreements and more focused on achieving desired outcomes.
Alan Nance from XLA Collab defines XLAs as putting the outcomes that determine the consumer experience at the centre of design.
Hannah Price from TOPdesk explains XLAs as how the customer feels about how they were treated.

That shift sounds simple. Operationalising it needs a different measurement infrastructure entirely, which is what experience management provides.

SLAs measure the process or the completion of an objective. XLAs measure the outcome and value.

What do XLAs look like?

XLAs have no universal formula. They should be tailored to your IT problem areas and to the outcome and value you want. That is what keeps them relevant in a way static SLAs rarely are.

You do not have to choose between XLAs and SLAs. Most mature IT organisations run both: SLAs governing operational processes, XLAs governing human outcomes. What matters is knowing which one tells you whether IT is working for people.

Unlike SLAs, XLA targets should evolve. Employee expectations change, services evolve, and what counts as good shifts over time. Build in a regular review cadence from day one.

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.

Not just sanctions. Rewards, too

One key difference with XLAs, and it leads to a different culture in IT, is rewarding performance rather than using the agreement to impose sanctions.

When your managed service provider knows their contract rewards measurable improvements in employee productivity, not just keeping SLAs green, the relationship changes. You are no longer managing against failure. You are collaborating toward impact.

Not just sanctions. Incentivise your service provider to make your employees more productive.

Sami Kallio, CEO, HappySignals

What XLAs do for the service desk

Measure the business value of the service desk
Increase co-operation between IT, providers, and the business
Motivate service desk teams with results they can see
Drive business value, not only process compliance

MSPs and the need for experience management

MSPs live and die by SLAs. But a contract that only measures process compliance gives your customer no visibility into whether their employees are happy, and gives you no early warning before renewal becomes a conversation.

The watermelon effect is real in MSP relationships. SLAs are green on the outside. Employees are unhappy. Nobody finds out until a dissatisfied customer does not renew.

Experience management surfaces how employees feel about the services they receive, in real time, before things escalate. Not a delayed CSAT survey or a quarterly review, but a continuous signal on what is working and what is not, across every customer. CSAT alone is a snapshot of sentiment, a yes or a no, not a why.

When both the MSP and the customer look at the same experience data, conversations shift from blame to collaboration. Issues get flagged and fixed faster. And because MSPs serve many customers, a fix for one friction point often resolves the same issue across ten others.

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.

Common mistakes when taking XLAs into use

XLA adoption is not always smooth. The mistakes fall into two groups: treating XLAs like SLAs, and the issues that come up during adoption.

Three ways to end up with a renamed SLA

PositioningBe honest about why your SLAs exist. If they are there because ITIL says so, or to defend IT to stakeholders, an XLA built for the same purpose gives you something to point to, not something to improve with. Every XLA should start with a question: what experience problem are we solving?
DesignSLAs are designed by IT, for IT. If the agreement is meant to reflect the employee experience, employees need to help define what good looks like. A service desk XLA for a logistics company looks different from one at a bank.
UseSetting the target, hitting it, and moving on is SLA behaviour. XLAs drive continuous improvement. The target is a floor, not a ceiling. Review it at least quarterly and let the data, not internal consensus, decide what good means.

Issues that come up during adoption

Confusion. The most common source of failure is treating an XLA as a renamed SLA. One word changes and everything else stays the same: the measurement culture, the governance, the review cadence. Define what experience means in your organisation before you launch. This is change management, not only measurement management.

Rip and replace. Do not rip out your SLAs. The organisations that succeed run both in parallel. Introduce XLAs as an addition that reveals what SLAs cannot see, not a replacement that removes what people already trust. Build a rhythm from the start: monthly check-ins on the data, quarterly reviews of the targets, an annual look at the XLA structure itself.

A number with nothing behind it. An XLA without an experience management infrastructure is just a number. To know why your score is where it is, which services drag it down, which employee groups are most affected, and what to do about it, you need continuous measurement, transparency, and an improvement cycle. XLAs are the target. Experience management is how you hit it.

Chapter 2

Why IT experience management is needed

Employees judge IT by the standards of consumer technology: instant, invisible, reliable. When IT falls short, they do not file a complaint. They lose time, lose trust, and route around IT where they can.

The cost is invisible in your SLA dashboard and very visible in your productivity data. In most large organisations, IT makes improvement decisions on ticket volume, escalations, and gut feeling. Those are lagging indicators at best.

Experience management gives IT a real-time signal on what is affecting people before it shows up in ticket volume, before it becomes an escalation, before it costs the business.

IT experience management is the practice of continuously measuring, sharing, and acting on experience data across all IT services, to improve outcomes for employees and show value to the business. It is how IT moves from keeping the lights on to enabling the business.

Experience is not just a new way of measuring. It is a cultural change.

Measure, share, identify, improve. Then measure again.

A continuous four-stage cycle for managing IT by outcomes rather than outputs. It works for any size of organisation, internal or outsourced, from scratch or on top of what you already measure.

Stage 1
Measure experiences
Start here, before you set a target. Without a baseline you are guessing where you are, and without continuous data you cannot tell whether improvements work. Most organisations are surprised by the first months.
Stage 2
Share the data
The step most organisations resist. Sharing low scores feels risky. What is risky is having bad data and not knowing it. When everyone sees the same picture, everyone works on the same improvements, and agents see the impact they already have.
Stage 3
Identify improvements
Without data, the agenda is set by whoever makes the most noise. Experience data shows where employees lose time and what drives the frustration: which services, which assignment groups, which ticket types close fast but leave people stuck.
Stage 4
Improve what matters most
The improvements that matter save employees time and reduce frustration, not the ones that make IT metrics look good. When they show up in the data, publish the before and after. Show the team what their work saved.

Organisations that have run this cycle for two or more years do not only have better scores. They have a different relationship between IT and the business.

Digital transformation and experience management

Digital transformation projects have a well-documented failure rate. The reason is almost never technical. Organisations measure delivery rather than adoption and impact. A system that goes live on schedule and sits unused is a delivery success and a business failure.

Experience management anchors transformation to the outcome that matters: did this change make employees more productive and less frustrated? That question can be answered continuously, with data, throughout the project.

Three ways it strengthens a transformation

MeasurementSLAs tell you whether IT delivered. XLAs tell you whether the delivery worked for employees. Only one tells you whether the transformation transformed anything.
DecisionsReal-time feedback during rollout is the early warning that catches adoption problems before they become business problems. Course-correct during rollout, not after the post-mortem.
CultureTechnology gets deployed and behaviour does not change. Experience data makes culture change measurable, and rewards in your XLAs align your providers with the behaviour you want.

Chapter 3

How IT can implement experience management

Experience management is not a project with an end date. It is a capability you build in layers. The organisations that do it best start simple and add sophistication over time.

Build it in this order

Continuous measurementBefore targets, before improvement programmes, before you talk to your MSP about XLAs: measure. Capture feedback at the point of each IT interaction, across the major service areas. The feedback that is not positive is where your roadmap lives.
TransparencyMake the data visible to service desk agents, service owners, providers, and business stakeholders, not only IT leadership. It turns a monitoring tool into a collaboration tool.
Experience targetsWith data and alignment in place, set targets that reflect what employees care about: happiness and productivity. Ambitious, grounded in your baseline, and reviewed as performance improves.

It is not your average CSAT questionnaire

CSAT and experience measurement are not the same thing, and confusing them is one of the most common reasons IT underinvests in the right data. CSAT asks: were you satisfied with that interaction? Experience measurement asks: did IT enable you to do your job?

The second question is harder to answer and much more valuable. It captures lost time, recurring friction, and the cumulative effect of IT on how people experience their work.

The HappySignals survey: one question per card, options that name the fix.

How to measure

The annual IT satisfaction survey is one of the most expensive things IT does, with little return. It is slow and generic, and by the time results are in, half the problems have resolved themselves or got worse.

The alternative: short, contextual questions at the moment of each IT interaction, tied to the service or ticket just resolved. Response rates of 15 to 40% against a typical CSAT baseline of 5 to 10%. A real-time signal instead of a snapshot. Evidence instead of gut feeling.

You do not need to survey everyone at once. Start with high-traffic touchpoints: service desk interactions, common incident types, recently deployed services. Build from there.

What to measure

Measure across all IT touchpoints, not only the service desk: devices, enterprise applications, remote working, onboarding. Each area has its own experience profile and its own owner. Give each owner their own view of the data.

Start narrow, with two or three high-volume service areas. Build the habit, show the value, and expand. Organisations that try to measure everything from day one end up measuring nothing well.

Setting experience level targets

Targets only drive behaviour if they are visible. Surface them in operational dashboards, in team meetings, and in the views your service owners and vendors use every day. An XLA buried in a contract nobody reads is not a management tool.

Service desk agents who see live experience scores respond differently from those who hear about results in a monthly review. Transparency at the operational level turns XLA targets into daily behaviour.

Treat targets as living documents. As performance improves and expectations rise, what counts as good will shift. Review at least quarterly, or you will hit the number and stop improving.

Setting a target in HappySignals: one score, one service, measured against today.

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.

13%

of tickets, the ones bounced between teams, cause 80% of lost productivity

HappySignals 2026 Global IT Experience Benchmark Report

26%

of lost employee time recovered in the first year

HappySignals customer average

15 to 40%

response rates, against 5 to 10% for a typical CSAT survey

HappySignals customer average

They did not just set XLAs. They built the practice behind them.

Two organisations that put experience management in place first, so the targets meant something.

Higher education, 22,000 staff and students

4h 56m3h 22m

perceived lost time per incident

Killing the watermelon: how ANU built one shared view of IT experience

IT is usually remembered for what breaks. With this, we can finally celebrate what works.Feisar Joya, Manager, Service Management Office, ANU

Before HappySignals, the Australian National University measured satisfaction with NPS tools that looked fine on the surface and missed what was underneath. “We used to have the watermelon effect,” says Feisar Joya, Manager of the Service Management Office: green on the outside, red at the core.

ANU started with culture, not metrics. Feedback across IT Services was reframed as a tool for improvement, not blame. Then they acted on the signal: colour-coded outage communications, restructured resolver groups to cut reassignments, and a follow-up for every piece of feedback scored below 6.

Crucially, NPS was never allowed to become a team scorecard.

Incident NPS rose to 83
Request NPS climbed to 88, lost time nearly halved
A follow-up for every piece of feedback scored below 6

Food and beverage, 275,000 employees

80%of global markets, without adding headcount

Scaling human-centric IT across Nestlé

Nestlé's IT has a clear goal. We want to be recognised as a business partner in our organisation.Osvaldo Santos, Senior Global Product Manager, Nestlé

Nestlé's IT had a clear ambition, in the words of Osvaldo Santos, Senior Global Product Manager: to be recognised as a business partner, not a back-office function. That meant an IT experience on par with the best consumer technology, and proof of value beyond process metrics.

Nestlé secured executive alignment first, with measurable goals for IT experience. It created a global, cross-functional Experience Management Office to coordinate efforts, share best practice, and localise improvements region by region, and shifted from project to product management so services keep evolving with employee needs.

Underpinning it all was a shift from project management to product management: IT services that keep evolving with employee needs.

A global, cross-functional Experience Management Office
Wi-Fi issues for a Northeast Asian team fixed from employee feedback
Internal benchmarking turned local wins into shared playbooks

After you book

From this guide to your first XLA.

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

Day 1
A 30-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 guide as a PDF.

The whole guide, formatted for reading offline and sharing with your team.

SLAs versus XLAs, side by side
The four stages of experience management
The mistakes that derail most XLA programs
Two customer cases with the numbers

Download the guide

By submitting you agree to HappySignals storing and processing your details to send you the guide and related content. See our Privacy Notice.