Defining activation, then walking every new business to it
New businesses signed up, but there was no clear line between "tried the app" and "started using it". I found that line in Amplitude, then built nudges that hand-hold each new business through the steps that cross it. An A/B test decided whether it shipped.
- My role
- Product + design. Defined the metric, designed and ran the nudges, read the test.
- Tools
- Amplitude for the data, Netcore for the nudges
- Metric
- Monthly activation rate
- Result
- 38% → 54%monthly activation, +42% relative
- How measured
- A/B test: no nudges vs nudges, run at the same time, then rolled out to all
The question
What does a business that sticks do in its first two days?
There was no shared definition of an activated business. Without one, we couldn't measure whether a first-week change worked, or point new users at anything specific.
So I went to the usage data in Amplitude and compared the businesses that stayed active with the ones that left. One pattern separated them: what they did in the first 48 hours after creating their business.
The definition we adopted
within 48 hours of creating the business
It is something a new business can do in two days, and the businesses that did it were the ones still active later. That made it a target we could design for and a number we could test against.
The idea
Hand-hold new businesses along the path active ones already take
The data gave us the target: 10+ items and 3–5 sales within 48 hours. Left alone, only 38% of new businesses activated in a month. The idea was to walk them through the critical features in that order, with nudges, so they saw what the app could do for their business and how to use it.
Before rolling it out, we set it up as a test against the app as it was:
No nudges
The existing app flow. New businesses find their own way.
Nudges along the path
Each new business is guided through the critical features, step by step, towards activation.
Both groups ran at the same time. If the nudge group activated more, nudges go to every new business.
The nudge journey
Nudges along the path to activation
Nudges, sent through Netcore, pointed each new business to the next step towards activation.
Business created
The 48-hour window starts.
Add 10+ items
Enough of their own products to bill for real, not a demo.
First sale
Record a real sale with those items.
3–5 sales
The business counts as activated.
Result
The nudge group activated more, so nudges went to everyone
Why it worked: businesses in the nudge group had been through every critical feature with a guide, so they understood what the app could do for them and how to use it. The nudges followed the path we'd found in the data, and the test showed the path was real.