3 Ways to Measure Team Performance (and What Each One Misses)

Your team's KPIs are green and the dashboard looks good. Then your best engineer quits, and the exit interview surfaces six months of problems you had no signal on.

This is the trap with how most teams measure team performance: the easy things to measure aren't always the things that matter. By the time the real issues show up in the data, they've already cost you.

So how do you know how a team is actually doing? There are three main ways to measure team performance: KPIs and quantitative metrics, qualitative measurement, and structured team assessments. Each one tells you something real, and each has a blind spot.

Method 1: KPIs and quantitative metrics

What is a KPI, and why do teams use them?

Key Performance Indicators (KPIs) are measurable values intended to show how effectively a person, team, or organization is doing towards specific objectives. They're your anchor measures to check in on to know if you're on the right path toward a goal.

For example...

• A support team tracking the degree to which customers felt helped after a customer service interaction (CSAT)
• A sales team tracking what percentage of qualified opportunities convert
• A marketing team tracking the number of leads who have shown enough interest to be considered a marketing qualified lead

Watching these values and how they change over time can teach you what to keep doing, what to stop doing, and how indicators influence each other.

How to set KPIs that mean something

To set meaningful KPIs, first consider the purpose of the team in context with the strategic goals of the organization. How does this team help the company be successful? Measuring an engineering team by the number of product sales doesn't make sense - they aren't driving sales. Consider what the team is driving and how you might measure how successful they are at that.

Goodhart's Law: when the metric becomes the target

When a measure becomes a target, it ceases to be a good measure.

The phrasing is commonly attributed to Goodhart, but it's actually anthropologist Marilyn Strathern's 1997 paraphrase of his 1975 original.

In practice, this means people may be incentivized to optimize for what's counted and quietly abandon what isn't (read: perverse incentives).

A recent Financial Times report revealed Amazon employees gaming internal AI usage metrics by "tokenmaxxing" — automating unnecessary tasks to inflate their token consumption and climb leaderboards. Amazon had set a target requiring 80% of developers to use AI weekly, and the metric did what measured metrics can unintentionally do: it got hit, just not in the way anyone wanted.

This is why measuring team performance well takes more than one lens. Here are two that work alongside KPIs without inheriting their failure modes.

Method 2: Qualitative measurement

Qualitative data is gathered through observation of first and second hand experiences. When KPIs are going up but something feels off, qualitative measurement might pinpoint what's going on.

Qualitative comes from Latin qualitas, interpretable as "of what kind." It's the sibling to quantus which asks 'how much/ how big.' So when measuring qualitatively, you're measuring what kind of team you have, rather than how much they produced.

Questions that surface qualitative data

By gathering first and second hand accounts that answer questions like these, you'll get a good idea of what the team is like. More often than not, you'll get ideas for how to improve the team based on the observations that answer these questions.

• When someone disagrees, what usually happens?
• How does this team decide what to work on next?
• What happens when something goes wrong?
• How does the team ask for help?

Anonymous check-ins as a scalable tool

Many companies use engagement surveys, or their quicker cousin pulse surveys, to measure how engaged and satisfied employees are. Typically these are used to quantify the things they are trying to measure, though some may include free-text components providing qualitative insights.

Many tools frame their questions as engagement and satisfaction with the company as a whole, while the data that comes back might really be more indicative of the team they spend their day-to-day with.

We created Team Pulse to quickly and anonymously uncover the qualitative conditions of teams that were previously invisible at scale.

The catch with qualitative data

Qualitative data can be harder than quantitative to decide what to do with, especially as a team where everyone needs to be on the same page about what the data even means. For more on what to do with qualitative data, our founder Joe Slatter writes about what to do when team alignment work has opened a can of worms.

Method 3: Structured team assessments

Many teams, especially in larger organizations, undergo structured team assessments.

What makes an assessment "structured"

Structured assessments are a defined instrument. They are carefully designed and tested over time to be repeatable, comparable, and produce a score that stays meaningful to track over time.

Validated instruments

The average "team assessment" you'll find are consultant-built and not peer-reviewed. Validated instruments, by contrast, are statistically verified, often in a university setting, to ensure they are measuring the things that make the difference. They also compare your team to normative data so you know how a team performs compared to "most teams" out there.

One of the most-cited validated team assessments is Amy Edmondson's Team Psychological Safety Scale. That instrument measures whether people feel safe taking interpersonal risks within the team.

At Better Practice we offer a comprehensive team performance instrument, the Benchmark, which measures your team in the areas of trust, clarity of purpose, team process, and orientation to conflict.

Stakeholder-derived scores

Stakeholder-derived scores come from those directly impacted by the work being done - often your customers.

One of the most common stakeholder-derived scores is the Net Promoter Score (NPS) - that question you see when a support chat closes and asks "On a scale of 0-10, how likely are you to recommend [thing] to others?" Fun fact: NPS was first introduced in a 2003 Harvard Business Review article called “The One Number You Need to Grow.”

Stakeholder-derived scores are useful because they give you data directly from those most impacted by the work being done. In other words, you hear how a team is doing from those most directly impacted by the work the team does.

The downsides to structured assessment

Validated instruments are thorough and accurate, but often involve surveys and take more effort from everyone on the team. They're also typically designed to be used infrequently. If everyone on the team is filling out the same survey questions every month, the results become less valid as they start measuring how good the team is at filling out surveys, rather than the performance of the team itself.

Stakeholder-derived scores, which are oftentimes also KPIs, have the same shortcomings as KPIs - they measure closer to the bottom line and leave you blind to the inner workings of the team.

Measurement only matters if you act on it 

In the end, measurement only matters if you do something with what you find. A weekly check-in nobody discusses is worse than no check-in at all, because the team has learned that their input doesn't go anywhere. Done well, regular assessment stops feeling like an audit and starts feeling like how the team gets better.

If you're interested in trying a Team Pulse for actionable qualitative insights, or measuring your team's performance with our university-validated instrument, check out the Better Together platform.