It’s a phrase I often repeat, but had never actually taken the time to write down.

Just yesterday, I was discussing business models with a client. And as often happens, the takeaway was simple: a well-designed company naturally generates the right behaviors — from both its employees and its partners.

Many organizations have lofty ambitions: performance, product excellence, positive impact… But without a deep understanding of how they actually function, they implement metrics that end up hindering those very goals instead of supporting them. The result? A disconnect between stated intentions and real-world behaviors.

Over the years, I’ve observed, unpacked (and sometimes helped fix) this kind of disconnect with several clients. Today, I’d like to share a few guiding principles to help you reflect on your own metrics — the ones you track, and the ones you impose on others.

Because a poorly designed KPI isn’t just a meaningless number. It’s a behavior you’re encouraging — whether you realize it or not.

So, what exactly is a KPI?

In English, KPI stands for Key Performance Indicator.

KPIs are measurable indicators.

For example:

• In customer service: average response time
• In marketing: campaign conversion rate
• In HR: employee retention rate
• In manufacturing: number of units shipped per day
• In customer experience (CX): customer satisfaction level

There’s an overwhelming number of things you can measure.

But a word of caution: not all metrics are created equal — and none of them tell the full story.

Let’s revisit those examples.

In customer service:

A low average response time might seem like a good thing. You might assume it’s due to an incredibly efficient call-handling team.

But in a real-world case, we found that in response to pressure to reduce response times, agents were transferring calls too quickly — passing customers off to other departments without really solving their problems. Result: each call was shorter, but customers had to wait longer between calls.

In the end, that tanked brand perception metrics… for a completely different team.

Oops.

In manufacturing:

The number of units shipped per day sounds great! The more you produce, the more you can sell. Perfect, right?

But in another case, we saw conflicting KPIs at play.

Imagine one person’s bonus is tied to daily production output. That person will do everything they can to go faster. Go go go!

Now imagine someone else is responsible for the quality of what comes off the line — minimizing defects, returns, and lawsuits.

Slow down!

The result? Constant tension about how to do the work.

Each person is acting based on their KPIs, motivated by their own incentive system.

Behaviors by design

The value and subtlety of business design is that it helps companies uncover problems that are often systemic — or even holistic.

Once the real issue is identified, that’s when the design work begins.

In the examples above, each metric is valid, and tracking them makes sense.

But you need to understand how to use KPIs — and which ones to show to which stakeholders. You also need to define which KPIs take precedence over others.

If your company says it’s committed to quality, then customer satisfaction, defect rates, and returns should matter morethan how much you produce each day.

If your company’s strategy is to be the cheapest on the market, then daily output and process optimization might matter more than quality or R&D investment.

These are just very simple examples — there are as many scenarios as there are businesses.

It’s also important to understand that there’s no one metric “to rule them all.” It’s about balance.

Of course, a company should always strive to improve across the board. But when it comes time to make decisions about desired behaviors… you need to understand what your metrics are actually driving.

Because designing KPIs means designing behaviors.

Still not convinced? Let’s talk about the snake problem.

In the 1960s, in some southern U.S. states, authorities wanted to reduce the rattlesnake population. So they created a bounty program: citizens were paid cash for each dead snake they brought in. It was a great metric — the more snakes killed, the greater the program’s “success.”

At first, it worked. People were killing snakes, and their numbers dropped.

But soon, some people saw a business opportunity:

They started breeding rattlesnakes on farms just to kill them and collect the reward.

When you reward a metric, people will eventually act in their own interest.

So ask yourself: What behaviors do you encourage when you introduce…

• A sales bonus?
• A billing rate per employee?
• A target for number of support tickets closed?
• A count of awards won per year?
• A view count on social media?

So how do you adjust your KPIs?

Every business is unique. KPI adjustment should be a cyclical exercise. Economies shift, companies evolve, people change. To stay competitive, a company must constantly reassess itself.

And the little exercise I suggest here is really just… a conversation. It’s called the Even Over strategy.

Concrete examples:

• Quality even over speed — We prioritize quality, even if it takes more time.
• User experience even over short-term profitability — We invest in UX, even if it doesn’t pay off right away.
• Transparency even over brand control — We choose honesty and authenticity, even if it’s risky for our reputation.

This kind of conversation helps companies identify which metrics align with their strategy — and how each team can influence what truly matters.

It’s about framing these as business metrics, not team or individual metrics. Once you establish that alignment, you can cascade down: define what to measure per team, then per person.

---

KPIs: a reflection of your true identity

These are choices that, if you design your company consciously, should reflect its identity. Identify who you want to be as a company — and design KPIs that encourage the behaviors you want to see.

If you say you’re all about quality, then prioritize and measure the quality of what’s produced — and reward people accordingly. Broadcast that commitment to quality, proudly. Don’t leave employees guessing how to ramp up production. Don’t pressure them on speed. Don’t evaluate them based on billing rates or the number of tickets they close.

If you’re an innovative company, measure the number of experiments and prototypes. Measure learning. Measure risk vs. potential. Don’t focus on KPIs for success or ROI. Innovation, by definition, is about the unknown.

If you’re a user-centric company, measure how often non-UX employees talk to users each month. Measure how often you validate user feedback. Measure how many actions genuinely bring your company closer to its users.

If you say you’re people-first and care about your employees, measure the number of activities held each month. Measure participation. Measure employee retention. Measure volunteer involvement in the community.

But above all… show up.

Not everything should be measured, after all. 😉