Do I need hundreds of thousands of users to see results with Aampe?
The question of whether an app has a large enough user population for us to learn about them is one that we get pretty frequently.
Indeed, most early stage founders and marketing/growth leads seem to think that high-quality inference and optimization requires “big data” — maybe hundreds of thousands or millions of users.
Can you really optimize push notifications when you’re only messaging a few thousand customers a month?
Yes. You absolutely can.
You don’t need tens of thousands of customers to start learning about them or optimally communicating with them. You can start with thousands, in some cases even hundreds.
In the study, they show how lots of data might even mislead you — something called the Big Data Paradox.
The paper’s authors looked at the accuracy of 3 surveys in estimating US COVID-19 vaccine uptake:
- One survey had about 250,000 responses per week
- The second had about 75,000 responses every two weeks.
The first survey overestimated the conversion rate by 17%, and the second by 14%.
Meanwhile the third survey only had about 1,000 responses per week, but it still provided reliable estimates and uncertainty quantification!
Here’s how the authors put it:
“We show how a survey of 250,000 respondents can produce an estimate of the population mean that is no more accurate than an estimate from a simple random sample of size 10. Our central message is that data quality matters more than data quantity, and that compensating the former with the latter is a mathematically provable losing proposition.”
Drive personalization with push
So, yeah, you can absolutely learn key things about your customers through your push notifications. In fact, they're uniquely suited for this purpose: Notifications are basically a beautifully scalable and highly targetable survey tool. They often have a singular focus, and they're delivered directly to your user's home screen.
Indeed, the question isn't 'can you do it...' It's 'why aren't you doing it yet?'