Startup Metrics in Plain English

Flow Ventures | January 10th

It’s a positive development that startups have figured out that metrics need to be at the core of their business and their pitch. Thanks to the Lean Startup, Dave McClure’sStartup Metrics for Pirates” and investors who are asking for a dose of proof with your passion.

I find that startup founders are more at ease with acquisition funnels, the viral coefficient, and cohort analysis. But many are getting lost in the weeds and losing sight of the big picture. You have a 3D cohort analysis graph (you know who you are…) but I have no idea what it means.

When you’re launching a new product, I think all of your key metrics can be derived from asking three simple questions:

  1. What is your core value proposition?
  2. How do you know people care?
  3. What’s the proof you’re delivering on your value proposition?

A shockingly large number of people still can’t define their value proposition in simple terms. E.g. we do A for B. The problem is, if you can’t even describe the core promise of your business, you can’t focus your product development, or market effectively, or measure your performance.

Customer acquisition is the time to test the promise of your business before actually having to deliver anything. This is where the fake “Buy” button works. If no one clicks on it, you don’t need to build anything. If your Facebook ads get no click throughs and no one makes it through your sign-up form, that’s the market telling you they don’t want what you’re promising and they don’t care if you can deliver it.

“Once I build my product I’ll be able to prove that customers want it.”

          – misguided entrepreneur

If you’re able to acquire customers that’s great news. But now you need to create metrics that prove that users are engaging in your product in a way that demonstrates value creation. This could be daily active use, amount of user-generated content, referrals to other friends or, obviously, spending money.

But you need to avoid the temptation to create vanity metrics that paint a rosy picture. You can’t build a business on 100k tire kickers from TechCrunch. But if you can find a few users that are truly engaged and truly getting value, you can probably find more of them. Make sure you set a high bar for what constitutes an “active user”. It doesn’t jive to say you’re disrupting an industry while making active user = “logs in at least once per week”.

Many products have more than one type of user. Not just “average users” and “whales” but people who derive different types of value from your product. In a marketplace product (real estate for example) you have buyers, sellers and brokers. All define value differently and need to be measured differently. The point is, you’ll probably have more than one metric that constitutes proof that you’re creating value overall.

Some Plain English Metrics

First, write down a 1-2 sentence value proposition. Seriously, stop avoiding it and do it.

  1. What acquisition metrics indicate a positive reception to your value proposition? Eg. effectiveness of paid and organic users; virality; activation rate.
  2. What is your definition of an “active user” and does this absolutely prove that you’re delivering on your value proposition? More clicks can be due to high engagement or bad UX… This is the toughest metric to design.
  3. Are engaged users maintaining or increasing their engagement over time? If not, how come?
  4. What % of acquired users never become active? Why?
  5. What % of engaged users drop-off? Why?

The most difficult metric to gather is why people stop using your product. By definition, these people are hard to talk to. Bend over backwards to talk to these people: offer them incentives or a personal email from the CEO or a compromising photo of the CEO. The data you get will be qualitative but you’ll be able to spot trends and make changes.

Answering the above five questions isn’t easy. One word of advice is not to worry about getting real-time data (you don’t need it) or perfectly accurate data (which you can’t get). You’ll probably have to throw in some qualitative data and wild guesses. That’s ok because at the beginning you’re looking for big obvious things. You’ll have plenty of time to optimize later.

Also, it’s expected that many of your metrics will suck. You’ll be trending down, not up. This is information you can use to change, fix, and pivot your way to success, or at least the next release.


Get back to basics by defining some plain English metrics for your business. If they’re well designed and information gathering isn’t crazily difficult, you’ll not only have a better view of your business but you’ll find it much easier to create meaningful projections. You’ll be able to have more intelligent conversations with your team and your investors, which hopefully are also taking place in plain English.