Most businesses track too many metrics and act on too few. The right KPIs are the ones that tell you something specific about whether your strategy is working — and they vary by business. Here's how to pick them.
Most businesses track too many metrics and act on too few. Dashboards accumulate. Numbers get reported in weekly meetings. And somehow, despite all that data, decisions still get made the same way they were made before anyone set up the dashboard.
The problem usually isn’t a lack of data. It’s that the metrics being tracked weren’t chosen carefully, don’t connect clearly to the decisions that matter, and change too often to tell a coherent story. Numbers that don’t drive action aren’t KPIs. They’re statistics.
A key performance indicator worth tracking has three properties. It’s measurable, meaning you can get a reliable number consistently. It’s tied to a decision, meaning it tells you something you’d need to know to act. And it’s actionable, meaning someone on the team can actually do something in response to what it shows.
“Website traffic” fails the third test for most businesses. If traffic drops, what exactly do you do? The answer requires more information. “Organic traffic from search” is better. “Organic traffic conversion rate by landing page” is better still, because it points directly to what might need fixing.
This distinction gets skipped in most metric discussions and it’s the most important one. A lagging indicator tells you what already happened. Revenue last month is a lagging indicator. Customer churn last quarter is a lagging indicator. These numbers tell you the score. They don’t tell you how to change it.
A leading indicator predicts what’s coming. Pipeline value is a leading indicator for future revenue. Support ticket volume is a leading indicator for churn. New trial sign-ups are a leading indicator for paid conversions. You need both types, but you manage the business on leading indicators. If you’re only watching the score, you’re always reacting.
There’s no universal list of the right metrics, because the right metrics depend on the business model. But here’s a starting point by function.
For sales: conversion rate (the percentage of leads that become customers), average deal size, sales cycle length, and pipeline coverage ratio (pipeline value divided by revenue target) are the core four. Pipeline coverage tells you whether you have enough in the funnel to hit your number. Sales cycle length tells you how long deals take, which matters for forecasting. If you only track one, make it conversion rate.
For marketing: cost per acquisition, lead-to-customer rate by channel, and channel-specific conversion rates matter most. These force you to evaluate channels on outcome, not activity. Content engagement rate is useful if you’re running a content strategy, but only if you’ve defined what “engagement” means for your goals.
For operations: fulfillment time, error rate, cost per unit, and customer support resolution time. These tend to be the metrics most directly controllable by operations teams and most directly felt by customers.
For finance: gross margin, net margin, cash runway if you’re pre-profitability, and accounts receivable days. AR days is undertracked by small businesses. If your average customer takes 55 days to pay on net-30 terms, you have a cash flow problem hiding in a good-looking revenue number.
For customer success: churn rate, net promoter score, customer lifetime value, and retention rate. NPS gets criticized for being imprecise, and that’s fair, but it’s a useful directional signal and easy to collect. Churn rate is the single most important metric for any subscription business.
More than five metrics per function means nothing gets real attention. The purpose of a KPI is focus, not comprehensiveness. If everything is important, nothing is. Pick the three to five numbers that most directly indicate whether the function is healthy, assign an owner to each one, set a target, and review on a consistent schedule.
The metrics you’re not tracking in the KPI set aren’t being ignored forever. They’re available if you need to investigate a problem. They’re just not being reviewed every week.
A functional KPI scorecard has four columns: the metric, the owner, the target, and the current value. That’s it. You don’t need software. A shared spreadsheet reviewed in a weekly meeting is enough to start.
The target matters more than most businesses realize. A metric without a target is just a number. “Conversion rate is 3.2%” tells you something. “Conversion rate is 3.2% against a target of 4.5%” tells you something actionable. It says you have a gap, and someone is responsible for closing it.
Customer acquisition cost compared to lifetime value. CAC is how much you spend to acquire a new customer across all marketing and sales channels. LTV is how much revenue a customer generates over their full relationship with you.
If LTV is less than three times CAC, you have a math problem. The business is spending too much to acquire customers relative to what those customers are worth. This is one of the most common ways a business can look healthy on revenue while slowly bleeding out on unit economics. You can’t fix a ratio you’re not watching.
The value of KPIs isn’t in the metrics you choose on day one. It’s in the habit of reviewing them consistently and acting on what they show. A good scorecard reviewed weekly is worth ten times a perfect one reviewed quarterly.
If you’re not sure which metrics to track or your data isn’t telling a clear story, that’s a solvable problem. See how we approach it at what we do.
