This is Probably how to Control and Engineer User Growth.

If you’re trying to get people to pay attention to something, you should probably think like an epidemiologist turned upside-down. Your goal should be to, like an epidemiologist, establish baseline estimates of expected growth and decay, but unlike an epidemiologist, find practical ways to infect everyone. Using this approach more accurately maps to the problem at hand, and seems to provide shockingly effective results.

Over the past few months I’ve written a whole bunch about how to measure, control, and engineer user growth. Here’s the four points I’ve come out of that exploration with — each one of them have a longer explanation driving that conclusion. But to quickly condense it all, here’s the gist:

1. You should think about growth in terms of epidemiology models, but turned around on their heads — instead of minimizing infection, you’re trying to maximize infection.
2. Just like people invented herd immunity as a clever, politically feasible way to minimize infection, we can invent other methodologies to maximize infection in the context of marketing.
3. You can, probably, reliably get a clear measure of baseline user growth. Just like researchers want to know how contagious a flu is in order to know how many people should be immunized, you should want to know how contagious your thing is in order to know how much of a push you’ll need to actually make, or whether it’s even worth it.
4. We can get baseline measures relatively quickly, and then we can introduce ways to (again, flipping epidemiology upside down) engineer virality. We can measure gains against expected baseline values to know exactly how successful we are. And we can grow with smart, targeted approaches that treat growth like the phenomenon it actually is.

There’s lots of complications in actually measuring these things accurately when we dive back down into the vast richness of actual human life. Exogenous impacts, weird properties of user behavior, the particularities of your particular app and how people use it, and even what constitutes an active user are all very hard problems to fix. I think that they can probably be fixed, though, or at least estimated. And right now, you can sign up at our website — we will be getting in touch very soon with the best use cases for rolling this type of thing out into the world.