# \[Essay\] Working backwards to find an activation metric

**URL:** <https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293>\
**Category:** Second Brain for Teams\
**Created:** [December 6, 2021, 3:48pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293 "2021-12-06T15:48:03Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![antoniokov](https://sea2.discourse-cdn.com/flex020/user_avatar/community.fibery.io/antoniokov/32/201_2.png) [@antoniokov](https://community.fibery.io/u/antoniokov)\
**Post date:** [December 6, 2021, 3:48pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293/1 "2021-12-06T15:48:03Z")

</div>

Finding an objectively good activation event for Fibery turned out to be pretty tricky. Two weeks of sweat and swears later, here is our ~~postmortem~~ practical guide:

> **[Working backwards to find an activation metric](https://fibery.io/blog/activation-metric/)**
>
> How we tried to pick an objectively good activation event for our non-linear product and invented something else instead.

What is the activation moment for your product? How did you find and validate it?

---

<div class="post-metadata">

**Author:** ![Bruce\_Onder](https://sea2.discourse-cdn.com/flex020/user_avatar/community.fibery.io/bruce_onder/32/272_2.png) [@Bruce\_Onder](https://community.fibery.io/u/Bruce_Onder)\
**Post date:** [December 7, 2021, 6:50pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293/2 "2021-12-07T18:50:43Z")

</div>

Great work! Am I missing something, though? Can you invite collaborators without having a paid account already?

---

<div class="post-metadata">

**Author:** ![mdubakov](https://sea2.discourse-cdn.com/flex020/user_avatar/community.fibery.io/mdubakov/32/10_2.png) [@mdubakov](https://community.fibery.io/u/mdubakov)\
**Post date:** [December 7, 2021, 6:52pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293/3 "2021-12-07T18:52:08Z")

</div>

Hmm, you should be able to do that in 30-days trial mode. If the trial is expired, then it is not possible

---

<div class="post-metadata">

**Author:** ![BanAnna](https://sea2.discourse-cdn.com/flex020/user_avatar/community.fibery.io/bananna/32/4654_2.png) [@BanAnna](https://community.fibery.io/u/BanAnna)\
**Post date:** [January 24, 2022, 4:36pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293/4 "2022-01-24T16:36:39Z")

</div>

Hi there! Just curious - how did you select the (relevant) events for the point system? Was that trial error or rather a statistical approach? Trying to solve a similar problem over here! 🙂

---

<div class="post-metadata">

**Author:** ![antoniokov](https://sea2.discourse-cdn.com/flex020/user_avatar/community.fibery.io/antoniokov/32/201_2.png) [@antoniokov](https://community.fibery.io/u/antoniokov)\
**Post date:** [January 24, 2022, 5:12pm UTC](https://community.fibery.io/t/essay-working-backwards-to-find-an-activation-metric/2293/5 "2022-01-24T17:12:53Z")

</div>

We simply pick the events subjectively based on the post-onboarding customer interviews, basic product analytics, and our understanding of the learning process within Fibery. No solid scientific method behind this so far 😅

This might have been pretty risky, if we wouldn’t have checked for the predictive power. Since we see that the score-based metrics predict the future non-score-based events (collaboration, purchase) alright, I sleep better at night.

However, there’s certainly [a huge] room for improvement — would love to discover your approach to this problem. Will send you a Calendly link via DM here.