Update cohorts.md #1516
Open
Update cohorts.md #1516
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Can I recommend that we remove the following section section for now? If we actually create a cohort with a "then" condition - i.e. watch a comedy THEN subscribe, we can sort of answer the question at hand. The main differences between our cohorts and funnels today are that: (1) cohorts are backward looking. I.e. If I'm looking at Jan 1, and my watch a comedy is the past 30 days, I'm looking at the users who watched a video in December and then subscribed over that same 30 day period only. (2) funnels are better in that our conversion windows are a moving target, i.e. if we watch a comedy on Jan 1, I have the next 30 days until Jan 30 to convert. (3) But funnels aren't the right answer necessarily today... because of our weird unique conversion criteria. Given all these nuances, not sure this is something that we want to address here. ### Cohorts and Time We don't recommend using Cohorts in analysis when trying to understand cause -> effect relationships between user behavior. For example, suppose you're trying to understand whether watching more comedy videos drives users to subscribe. You might create a Comedy Fans cohort defined as "Users who performed Watched Video where genre=comedy in the last 30 days" and then breakdown your Subscriptions by whether users are in the Comedy Fans cohort. The problem is that the user might have become a Comedy Fan _after_ subscribing, which doesn't tell you whether watching comedy had any impact on subscribing. Instead, we recommend using Funnels if you want to understand specific sequences of events with a strict ordering between each other. For example, creating a funnel from Signup -> Subscription and comparing that to a funnel from Signup -> Watched Comedy -> Subscription, to understand what % of users convert after watching comedy.
PeishanFromMixpanel
requested review from
mavlee
and removed request for
a team
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Can I recommend that we update the following section section? ### Cohorts and Time
Cohorts and Time - update to:
A common question that customers want to answer is how certain user behaviour impact downstream behaviour. Should you use cohorts or funnels to answer the question of how many users watch a comedy video and then subscribe? Cohorts and filters will give you different answers.
For example, consider this report: https://mixpanel.com/s/4AuMRi
For the cohort on Jan 1, I'm looking at the users who watched a comedy video in the last 30 days before Dec 25 and who then subscribed over the same period of time. However, for the funnel on Dec 25, I'm looking at the users who watched a comedy video on Dec 25, and who have then also subscribed anytime within the next 30 day period from Dec 25. This means that if a user subsequently subscribes on Jan 3, the funnel chart can change to take that conversion into account. In this case, if you care about the conversion window, i.e. the amount of time that a user has in order to count as a conversion, using funnels would give you a more accurate answer.