5D7: the activation metric that cut churn 86% for the users who hit it
If a new customer did not make it past their first month, they almost never came back. We worked backwards from retention to find the moment that made people stay.
Lumen5Sr. Growth Marketing Manager & AdvisorDec 2019 – Nov 20223 min read- 86%
- Lower churn for users who hit 5D7
- 5 in 7
- Video downloads in the first 7 days
- 33%
- Churn in bad months, before
Context
Churn was Lumen5’s biggest problem, running as high as 33% in bad months. Looking at it month over month showed exactly where the damage happened: customers who did not stay past their first 30 days almost never retained at all.
That made the first month the most important part of the business. If we could find what separated customers who stayed from those who left, we could steer every new customer toward it.
The thesisRetention is decided early. Find the moment a customer first gets real value, define it precisely, and build the first week around reaching it.
The activation states
- 01Sign-upA new paying customer arrives with an intent, but has not experienced any value yet.
- 02SetupEverything they must do before the product can deliver: brand assets, first content, first draft.
- 03AhaThe first time they get real value: a finished video they can actually use.
- 04HabitMaking videos becomes part of how they work, week after week.
The approach follows the activation model taught at Reforge, which splits a new user’s journey into distinct states. Setup is everything a user has to do before the product can deliver value. The aha moment is when they first experience that value. The habit moment is when using the product becomes part of their routine.
Treating these as separate states matters because each one fails for different reasons and needs different fixes. A user stuck in setup needs less friction; a user who has had their aha moment but no habit needs reasons and reminders to come back.
Working backwards to the metric
Rather than guessing which features mattered, we started from the outcome we wanted, customers retained past the first month, and worked backwards. We compared the early behaviour of customers who stayed with those who churned and looked for the actions most strongly correlated with staying.
The final step was to turn that correlation into a precise target in the form “X actions within Y days”, the same shape as well-known activation metrics at companies like Facebook and Slack. A precise target is something a whole company can design around.
The metric: 5D7
The data pointed to one clear threshold: after buying any plan, a customer needed to download five videos within their first seven days. They could be five different videos, or the same video exported in five formats for different channels.
We called it 5D7, and it became the target for onboarding, lifecycle messaging, product and customer success. Every touch in a new customer’s first week was designed to move them toward it.
Downloads were the right signal because a download is the moment a video leaves Lumen5 and gets used. Five downloads meant the customer had either made several videos or adapted one video for several channels, such as their website, TikTok, Facebook and Instagram. Either way, Lumen5 had become part of their workflow.
Putting it to work
- Action-based marketingEvery message was triggered by what a customer did, or did not do, inside the app.
- In-app testsWe tested pop-ups, reminders and layouts to prompt the next download at the right moment.
- Email and promotionsLifecycle emails and offers followed up on in-app behaviour to pull customers back toward 5D7.
- Feedback to productWhat we learned about where customers stalled went straight to the product team, so the product itself made 5D7 easier to reach.
Customers who reached 5D7 churned 86% less, month over month, than customers who did not. Because the first 30 days decided almost everything, moving more customers past that threshold lifted the growth ceiling for the whole business.