Which numbers to track from your first day

A new project throws off a lot of numbers. Page views, downloads, Instagram followers, the WhatsApp status views from the day you announced it. Most of them feel good and tell you nothing about whether the product works.

The three below answer the questions that matter in the first months. Is anyone arriving? Do they stay? Do they get what the product promised? If you can answer those every week, you'll know what to work on next. Everything else can wait.

Count new sign-ups per week, not per day. Daily numbers jump around with payday, exam season and a bad network day, and they'll make you either panic or celebrate for no reason.

Then look at the rate. Paul Graham, who co-founded Y Combinator, wrote that a good growth rate during the programme is 5 to 7% a week, and 10% is exceptional. He also says the absolute number matters less than the ratio of new users to the ones you already have.¹

So 20 sign-ups in a week means something different at 100 users than at 2,000. Write the percentage next to the count each week. If you don't charge yet, he suggests tracking active users as the main number instead.¹

This is the number most builders skip, and the one that tells you the most. Group people by the week they signed up, then check how many of each group are still using the product a week later, a month later, three months later.

Google Analytics does this for you. Its retention report shows how many new users return each day over their first 42 days, with lines for day 1, day 7 and day 30, grouped by the day people arrived.²

What you're looking for is the shape. If the line falls and then flattens out, some people have found a reason to stay, and you can build on them. If it slides all the way to zero, more sign-ups won't fix it. Talk to the people who left before you spend a naira on marketing.

Every product has one action that shows a person got it. For an invoicing app, sending the first invoice. For a tutoring app, booking the first class. For a POS tool, recording the first sale. Pick yours and write it down.

Then count how many new sign-ups do it, and how long it takes them. Someone who signs up and never sends that invoice was never really a user. If only a third get there, the problem is often the first five minutes: a long form, an OTP that arrives ten minutes late, a screen that asks for too much before showing anything.

Track it as a named event. Firebase, Google's analytics for apps, is free with no limit on reports and handles up to 500 different events.³ You need maybe five. Call them plain names you'll understand in a year, like invoice_sent.

Graham's view is that revenue is the best number to grow, and active users are second best for products that don't charge yet.¹ Once money comes in, track it weekly next to the other three.

Keep it simple. Money received this week, number of people who paid, and how many paid again from the month before. If you sell subscriptions, the last one tells you more than the total does.

For a website or web app, Google Analytics is free. For a mobile app, Firebase does the same job and plugs into Google Analytics.³

PostHog is another option. Its free plan includes a million events and 5,000 session recordings a month, with no card needed.⁴ Watching a few recordings of real people getting stuck on your sign-up screen will teach you more than a week of guessing.

And if you'd rather not add another tool, a query on your own database every Monday works fine. Sign-ups by week, logins by week, the key action by week. Paste the answers into a sheet.

Total sign-ups since launch only ever goes up, so it can't warn you. App downloads count people who installed and never opened. Followers and page views measure attention, not use.

They're fine to mention in a pitch or a post. Just don't make decisions with them. The people who signed up last week and came back this week tell you far more.

On Prow: Posting your project on Prow is free. Your project page shows how many people tried it, who follows it and the feedback they leave, and you can post updates as you build. Once it earns, connect Stripe, or send Paystack or Flutterwave screenshots, and your revenue shows on the page too. If you ever sell, these are the numbers buyers will ask about first, so it helps to have tracked them from the start.

• You count sign-ups by week and write the growth rate next to them.
• You know how many people from each week are still around a month later.
• You've picked the one action that shows someone got the product, and you track it.
• Revenue goes in the same sheet the week you start charging.
• You use one free tool, or a Monday query, and nothing more.
• You ignore totals, downloads and followers when deciding what to build.