Startup KPIs to track at every stage

The startup KPIs to track depend on your stage. Before revenue, track activation, early retention, cash, burn, runway and how fast you learn. At seed, add revenue growth, revenue retention, early unit economics and sales efficiency. Getting ready for a Series A, track ARR, net revenue retention, CAC payback and gross margin.

A KPI you cannot act on is a number you report out of habit. The test for every metric below is the same: if it moved sharply next month, would you do something different? If the answer is no, it belongs in an appendix, not on your dashboard.

Why do startup KPIs change by stage?

Because the question the company has to answer changes. At pre-seed the question is whether anyone wants the thing. At seed it is whether you can sell it again and again. At Series A it is whether more capital turns into more revenue rather than just more burn. A dashboard built for the wrong question looks busy and tells you nothing.

The other reason is noise. Unit economics on eleven customers are mostly anecdote, and a pre-revenue team calculating CAC payback is doing arithmetic on a guess. Track what your data can actually support at your size, and add metrics as the sample grows.

What should you track at pre-seed?

Before revenue, the job is to learn fast without running out of money while you do it. Five numbers cover it:

• Activation. The share of new users who reach the first moment of real value, defined by one specific action. It decides what to fix in onboarding before you send anyone else into it.
• Early retention. Of the users who activated in a given week, how many are still active four and eight weeks later. It decides whether you have something to scale or something to rethink.
• Cash in bank. The closing balance from the bank statement, not from the model. It is the constraint every other decision on this list answers to.
• Net burn and runway. Monthly spend minus revenue, and cash divided by that figure. It decides when you start raising and which experiments you can afford first.
• Learning velocity. How many customer conversations, experiments or shipped iterations you complete each week. It tells you whether the team is testing assumptions or just building.

Write down what "active" means before you report on it. The a16z essay on 16 startup metrics points out that companies have almost unlimited definitions of an active user, and some charts never say which one they used. Pick one action, name it, and keep it. If you want the runway math worked through properly, read how to calculate startup runway.

What should you track at seed?

Once money is coming in, the question becomes repeatability. The pre-seed cash numbers stay on the dashboard, and these join them:

• Revenue growth rate. Month-over-month growth in revenue or MRR, with the trailing three months visible. It tells you whether the plan you raised on is still the plan.
• Revenue retention. Revenue from last month's customers that is still there this month, before any upsell. It tells you whether growth is compounding or leaking out of the bottom.
• Early unit economics. What a customer contributes after the direct cost of serving them, next to what it cost to win them. It drives pricing and which customer segment to chase.
• Paid and blended CAC, separately. Acquisition cost for customers you paid to reach, and for all new customers. It decides whether a channel deserves more money.
• Sales efficiency. Win rate, sales cycle length and average contract value from your pipeline. It decides whether to hire the first salesperson or keep selling yourself.

For a sense of scale on growth, Paul Graham's essay Startup = Growth puts a good growth rate during YC at 5 to 7% a week, measured on revenue, or on active users for startups that are not charging yet. On acquisition cost, the same a16z essay explains why investors weigh paid CAC above blended: blended CAC includes customers who arrived organically, so it cannot tell you whether paid acquisition scales profitably.

What should you track to be ready for Series A?

A Series A investor is buying evidence that the machine works and will keep working at a bigger size. These numbers need enough history to show a trend, which means you start tracking them a year before you raise, not a month:

• ARR. Annualized recurring revenue from contracts in force, excluding one-off and services revenue. It sets the size of the round and the plan the round funds.
• Net revenue retention. This year's revenue from last year's customers, including expansion and net of churn and downgrades, divided by what they paid a year ago. It shows how much you grow without a single new logo.
• CAC payback. The months of gross profit from a new customer needed to recover what it cost to acquire them. It decides how hard you can push acquisition spend before cash becomes the limit.
• Gross margin. Revenue minus the direct cost of delivering it, as a percentage of revenue. It drives pricing, hosting and services decisions, and how an investor values each dollar of ARR.
• Burn multiple. Net burn divided by net new ARR, the shortest summary of how efficiently cash turns into growth.

Report gross retention beside net retention. The a16z essay makes the point that net revenue churn understates your real losses because it blends upsells with churn, and an investor will separate the two whether you do or not.

How is this different from what you report to the board?

Tracking and reporting are two separate jobs. You track everything above, often weekly, because you are the one making the decisions it drives. The board gets a fixed, shorter set every month, and the post on which metrics to report to your board covers what goes in that set and why it should never change. Investors get shorter still: three to six numbers in a monthly investor update.

Your stage KPIs feed those reports. They do not replace them, and they should not all appear in them.

What makes a startup KPI dashboard useful?

Three things. Every number has a written definition that never changes quietly. Every number has a source, ideally a system rather than someone's memory of a spreadsheet. And every number sits next to last month and the plan, so you read a trend instead of a snapshot. A dashboard with forty tiles fails the first test within a quarter.

Boxsy's control center is built around those three rules. Every KPI carries a definition and a formula tooltip, so ARR means the same thing in March as it does in September. Revenue, bookings, ARR, cash, burn, runway, and sales metrics such as win rate and average contract value fill automatically from QuickBooks, HubSpot and Plaid, and anything you have not connected can be entered by hand or in a monthly check-in. See what is in each plan, or read how that data is handled and secured.

Fewer numbers, tracked for longer

The founders with the clearest view of their company are not the ones tracking the most metrics. They picked the handful that fit their stage, wrote down what each one means, and kept them steady long enough to see a trend. When the stage changes, change the set on purpose, and write down why.

If you are not sure which stage your numbers actually put you in, talk to our team and walk through them with us.