Something’s Not Adding Up

Ask a leadership team what single number they're optimizing for above everything else, and you'll probably get a pause before you get an answer.

That pause is telling.

You can see the result on plenty of company scorecards. Every function hits its targets, but somehow the business still isn't where it needs to be.

So what went wrong?

The problem is the scorecard itself. The company has a collection of metrics rather than a measurement system.

Metrics are the language of business, and when the metrics aren't clear, it's hard to plan or execute effectively. Teams have to know what they're working on and how it connects to what the organization cares about most. When those connections aren't clear, every function can feel successful in its own lane without the company actually succeeding.

Metrics already have a hierarchy

Most business metrics have a relationship to other metrics whether anyone draws that relationship out or not.

Some relationships are mathematical. Revenue might break down into traffic × conversion × average transaction value. Others are based on cause-and-effect assumptions. Improving product quality might increase retention, which increases customer lifetime value, which contributes to revenue and profit.

Either way, there should be a defensible path upward.

A spreadsheet often hides that hierarchy. Every metric appears at the same visual level, even though they don't carry the same weight.

A KPI tree makes the relationships visible. Think of it as an org chart for your numbers.

Draw those same metrics as a diagram, with lines connecting each number to what it's meant to move, and you can see pretty quickly whether a metric has a real path upward or is just sitting there in isolation.

One measurement system

Companies often separate operational KPIs from financial KPIs. That distinction can be useful, but they shouldn't live in separate systems.

There should be one hierarchy.

Some numbers happen to be measured in dollars. Others are percentages, units, customers, products, days, or hours. That's fine, as long as each has a stated connection to what's above it.

So many companies never build that hierarchy. Instead, every function develops its own pile of metrics and calls them KPIs.

But whose "key performance" are they measuring?

I've made this mistake myself

I've gotten this wrong more than once over the years.

I've been so focused on understanding exactly what my own initiative did that I lost track of understanding how much it moved the overall business.

It's an easy trap.

You look at a big number like revenue or margin, or a rate like conversion, and think that too many things touch it to isolate your own impact. You want to know what your team accomplished, so you create something narrower that you can control and attribute directly to your work.

The problem is that the easier a metric is to attribute entirely to your team, the further it can get from the result the business actually cares about.

The desire for attribution pushes leaders toward metrics they can control, while enterprise performance depends on outcomes no function controls alone.

When every function does this at once, you can end up with a scorecard full of green that doesn't add up to anything meaninful

Follow the number all the way through

Marketing is an easy example.

Attributing a sale directly to marketing can be hard. Pricing, product, timing, inventory, the sales conversation, and plenty of other things may play a role. So marketing might focus on something it can clearly control, like leads generated.

The problem is that leads can grow every month while revenue goes nowhere. The leads might be poor quality. Sales might not follow up effectively. Customers might enter the pipeline and never close.

The useful question is how the numbers connect:

Leads generated to qualified opportunities to pipeline value to closed revenue

Now you can see the system.

Marketing may influence the first few numbers more directly. Sales may influence the later ones. Neither function controls the entire chain.

That's precisely the point.

Functions and initiatives operate as one system, and what matters is what they contribute to it together. The handoffs between functions can matter as much as what happens inside them.

A performance metric should have a defensible path back to the number the business cares about most. The exceptions are things you've deliberately decided to protect regardless of that path, like safety, compliance, or another critical constraint.

Otherwise, it can provide a lot of confidence without telling you very much about whether you're winning.

Start with the number that actually matters

The number that could organize all of this often sits in finance or a GM's office, treated primarily as a financial statement rather than as the map the rest of the company's metrics should be drawn from.

That's a missed opportunity.

The P&L should be more than Finance's reporting instrument. It can provide the foundation for a shared operating model.

Start with the number the business ultimately cares about, then work backward.

There isn't one North Star number that makes sense for every organization. The right number depends on the economics and objectives of the business.

But you do have to choose one, and you have to be specific.

"Profit" isn't specific enough.

Operating income, net income, EBIT, and free cash flow are different measures that can encourage different decisions. Free cash flow pushes a team to think about things like collections and inventory. A profitable-growth target can encourage a team to walk away from revenue that doesn't carry enough margin. EBIT can look fine in a period when cash is deteriorating because working capital is moving in the wrong direction.

A company can perform well against one of those measures while struggling against another.

A simple test for whether your North Star is specific enough is to ask finance, sales, and operations to define it separately.

If you get three different answers, it isn't specific enough yet.

Build down before you build out

Once you've identified the North Star, start decomposing it.

What directly determines that number?

What determines those numbers?

Keep working your way down until teams can see the metrics their work actually influences.

Only then should you start inventing new metrics.

An initiative metric is temporary, tied to a specific effort with a start and an end. A functional metric is ongoing, tied to the regular work a team does every day. In either case, the first move should be to see whether an existing business metric can simply be carved out more precisely.

Pipeline value, for example, is a more specific cut of the path toward revenue. You haven't invented a new definition of success. You've made one part of the existing system measurable.

Don't create a new metric until you're really sure nothing existing captures the connection you need. And when you do create one, make sure its connection back to the North Star is clear.

Try this with your leadership team

Pull out your leadership team's scorecard.

Pick any KPI and ask the executive who owns it to trace exactly how moving that number contributes to your North Star.

Then pick another.

Keep going.

Some connections will be mathematical. Others will depend on assumptions about cause and effect. That's useful too, because now those assumptions are visible and you can test whether they're actually true.

You may also discover that members of the leadership team have completely different assumptions about what drives what. That's exactly the kind of disagreement you want to find.

And pay particular attention to the metrics everyone is hitting.

If every executive on your leadership team can hit their KPIs while the company misses its most important number, you have to ask a pretty uncomfortable question:

What exactly were those KPIs measuring?

If you can't draw the connections, you don't have a measurement system yet. You have a collection of numbers.

And that's how your whole scorecard can be green while the business is still losing.