B2B Startup Metrics with Tom Blomfield | Startup School

**Tom Blomfield** (0:09)
Hi there, my name is Tom Blomfield. I'm a Group Partner at Y Combinator. And today, we're gonna be talking about one of my favorite topics, metrics, and why they're so useful for startups. So why are metrics important? First of all, it's pretty obvious that with better metrics, you'll make better decisions.
It's like flying an aeroplane. With no instruments, you're flying blind. You don't know what's happening to the aircraft, and you're not in control. Having great metrics is like having great instruments in an aircraft, it lets you tweak and iterate and make sure you're really in control of your startup. We often see, and I've seen even in the last two or three weeks, founders who've had these great launches, and they've launched on Hacker News, on Product Hunt, and they've had hundreds of people come and use their service day after day, but they have no idea how many of those users are new users or returning users. They don't know if they're daily active or weekly active. They could be churning off all of their users instantly, and they don't know at all. So the first thing they do after launching blind is to go back and build metrics in. We would advise you don't do that. You should build basic metrics into your product before you launch.
As an investor, it's really easy to tell founders who are in command of their metrics versus founders who aren't. And it's really impressive when founders can talk about what percentage of their signups are DAU or WAU, or what the annual revenue per user is, and we'll go into some of these in detail, but it's a big differentiator when a founder can talk so fluently about these metrics. Before we dive in, I want to give a couple of warnings.
The other extreme is also bad.
So it's a founder who, even before they launched, has a dashboard with perhaps 500 metrics. Maybe they've been a product manager at a big tech company, or they've just watched too many YouTube videos, but they want to make every decision in their startup with metrics. And when you only have a few hundred or a few thousand users, that's basically impossible. They want to split test everything, should this button be blue or green. And frankly, it doesn't matter, and you don't have the volume of users or data to make those kind of split tests sensibly. So what you should do is certainly split test the really important decisions. Should the cost per user be $80 per year or $200 per year? That's a really good experiment to split test. But making buttons red or green, that's not really something you have the scale to split test until you're really at the size of Google or Facebook. A final warning, don't hide behind your metrics. You've still got to get out of the building and talk to customers.
Brian from Airbnb still hosts Airbnb users in his home. It's an obsession with staying close to customers, so you can't let metrics get in the way of that. So let's get started. You're planning a product launch in perhaps a week or two, and maybe you've not got any metrics in place yet. What do you do? The first thing is to pick four or five key metrics to track accurately. Not 30 or 50 Four or five is fine. This number will grow over time. We'll talk about what those key metrics should be in a little bit. You should pick the most straightforward analytics solution you can operate. It might just be your SQL database, making simple SQL queries to count to the number of signups. Post Hog from Winter 2020 has a great SQL analytics tool you can use on top of pretty much any SQL database. So you should check that out. You should also agree the definitions of these four or five key metrics and stick with them. So it might not be the absolute perfect definition of an active user, but constant arguments about what your key metrics are are even worse than having no metrics at all. So your whole team has to come together and agree that an active user is someone who uses the product every day, or at least once a week, or at least five times a week.
It honestly matters less the precise definition than you actually all agree with it.
I remember so many disagreements where the marketing team said, we've sent you 2, 500 new leads this month, and the sales team says, no, no, no, they weren't qualified leads. They don't meet our definitions. And this disagreement internally just destroys the productivity of meetings where metrics are involved. So you really have to have centralized definitions of metrics that are written down and everyone agrees on. So say you launched, and perhaps the metrics aren't quite what you hoped. The weekly active users aren't quite as high as you originally wanted. In that situation, founders are often tempted to pick a different metric or change the definition of those metrics. So instead of a weekly active, let's go for a monthly active. The number looks a bit better. Honestly, you're only fooling yourself in this situation. It's really, really important that you keep the definition of your metric consistent over time to see if you're improving or not. And that's why it's so hard to compare metrics between different companies.