@future-teacher-7046 @fresh-football-47124 Do you have any plans to provide more significant reports on the experiment ->result section? Let me give you an example, Lets say I am running my experiment with multiple metrics and I want to know actually which groups or sub-groups are doing particularly very Good or very Bad. The sub groups can be anything , Like age-range, country/region etc. I know that I can segregate the results using custom dimensions, But what I want is that GrowthBook to provide me those extra insights that I may have missed or didn't know.

Hello I have added my existing mysql as data source, should I create a table with the expected fields, like the sample below? Otherwise, how do I edit experiments query?

SELECT
  user_id,
  anonymous_id,
  received_at as timestamp,
  experiment_id,
  variation_id
FROM
  experiment_viewed

Hi, is there a way to change the default tracking key generated by GrowthBook?

Hi there. Is there an efficient way to run A/B test for increasing retention? I see there is an option to set up a metric of the "duration" type but it's not quite the same

Also, when connecting Google Analytics as a data source, there is no option to choose a specific Google Analytics account in case there are several, so not sure which account GrowhBook connects to

So there is a "duration" metric type with the example query

SELECT
  user_id as user_id,
  anonymous_id as anonymous_id,
  duration as value,
  received_at as timestamp
FROM
  sessions

What happens if there are several entries with the same user id? Do we sum all the values of the same user or calculate their average?

I am trying to set up a query for a metric but something got me confused. The text under Example SQL explains that value is the "number of conversions" for this user_id, but the "timestamp" is the "the time the action was performed". So what's the best way to store the metric data? Aggregated or raw? It's a revenue-based metric, then the "value" would represent the order sum

Hi all, I have a couple of question about

stopped

experiments : 1. When I stop an experiment flagging it as

Winner

, the remote config properly returns the property

force

for this stopped experiment, specifying the variation ID that won. Thanks to this, I can display the winning variation to all my users meanwhile I clean the code from this experiment. But for all other results (

Lost

,

Did not finish

,

Inconclusive

) there is not

force: O

on the CDN remote config. Is there a reason explaining that? 2. Why when an experiment is stopped,

groups

are removed from the remote config? Because while we clean the code after the experiment has stopped, we still want to deliver the appropriate variation only to the targeted groups. Despite the experiment has stopped.

Hello everyone! I am a beginner js developer using mixpanel for users activity data analytics. I am looking for a way to implement some A/B testing that is a bit more complicated than displaying versions of CTA on buttons (I am talking about displaying different versions of routes for different subsets of users for example). Is GrowthBook somehow helping with handling/running a/b experiments? Would you recommend some reading about how to properly go about this with growthBook beside growthBook's Experiments documentation chapter and also point me to some good code example? Thanks a lot

Hey, hey ! Do you have any guidelines about how to handle

development to production

AB testing workflow ? For instance, we are developing a feature, and want to setup an AB Test on our development platform to check that everything is setup correctly. Development AB tests are plugged to our Mixpanel Development workspace. When I want to go live, I must duplicate my experiment to make it target our production Mixpanel Workspace, but the side effect is that the webhook config will now reference twice my AB/test. Do you have recommendation about how to handle such a workflow? Thx!

can someone explain what

anonymous_id

is defined as?

I have a question that is somehow related to the one asked above. I wonder how do I make sure that same experiment variant is shown to the same user upon future sessions, and also to the same user after they signup? If I understand correctly this is what the

user: { id: "123" },

is for in the quick usage example here https://docs.growthbook.io/lib/js I am using mixpanel so I am thinking of somehow accessing its distinct_id as it should be automatically merged to id cluster after signup. However, i am not sure how to access it. It seems like bread and butter of experiments so I bet someone will have an idea how to go about it. Thanks in advance! 🙂

Hello 👋 I am wrong or the "Activation metric" disappeared from the cloud app? I don't find it anymore

Hi all 🙂 I don't see how to use Segments (https://app.growthbook.io/segments) on an experiment once it has been created? Thank you

Hi. The dashboard is showing a wrong experiment status.

Hi. The Experiment Results page is not showing the percent change bands properly, it is sometimes hiding part of the interval

Hi everyone. How do I see the estimates for experiment runtime? I've set up metrics and pageviews which, according to the docs, should give me this "This is used to analyze historical user behavior. For example, we can use this data to predict how long an experiment will take before you start." But I wasn't able to find this info anywhere. Thanks in advance!

Hi everyone, I have a problem connecting with the mysql data source. Specifically I receive the following error for Postgresql

connect ECONNREFUSED 127.0.0.1:5432

Fields overlap when testing more than 2 variations

Hi, I have a question on access to Growthbook, do all user need to be added as "team member" to be able to view what experiments are being run, view results etc? Thinking like stakeholders and consumer of the experiments who don't really need to create experiments or add metrics etc

Is the window.GROWTHBOOK_CONFIG definition on client side optional?

Is there any documentations I can read on segments? How should it be defined and used in experiements?

Hi. My understanding is that the statistical significance for non-binomial metrics is calculated somewhat different and uses different statistical approaches, is that correct?

I am seeing this "value changed too much" notification and while I understand the concept and the possible problem with the experiment premises and setup, I'd like to know how I go about this problem. After checking the experiment setup, everything looks fine and the tracking works OK. Should I wait for the "suspicious results" get less suspicious or is it the final verdict? Looks like our user behaviour is rather different depending on the day of week, so maybe that's that. Should I wait for more data?

I am having trouble uploading screenshots for variations, is there any tips around it like image size etc?

Hi, I have difficulties to setup the first experiment with visual editor Background: I use Matomo Analytics as a Wordpress plugin, so the data source is MySQL. The Metrics has been set up already, it displays the results. I created an Experiment table with the suggested fields within the MySQL DB. But the Experiment script does not work. Should I write an extra module to put the data into this table?

track: function (experimentId, variationId) {
      analytics.track("Experiment Viewed", {
        experimentId,
        variationId,
      });
    },

This message was deleted.

Hi everyone. I'm checking the underlying queries that calculate the experiment results and have a question. It looks like the conversion attributes to the earliest pageview found for a particular user_id, in which case this scenario applies in case of the conversion window being 24 hours.

1.user views a test variation 
2. user views a test variation again after 3 hours
+ user converts after 27 hours after viewing the variation for the first time.

conversion doesn't get attributed to the variation.

Just want to check my assumption at the time, but it looks like it makes sense for the conversion window to get updated each time the user views the variation (although I am not an expert in a/b tests)

Hey everyone, sorry in advance this may be a real beginner question. I need to track multiple experiments and have the ability to dynamically assign users/clients/records/etc. to a group without experiment overlapping. e.g. I want to ensure a User in the test group in Experiment A is not in the test group for Experiment B. Is GrowthBook capable of that? Also, I see the data connectors to Athena, Redshift, BigQuery, and other querying tools. Is GrowthBook aimed at experiment assignment live? I'm wondering how I could send an HTTP/Webhook request to my GrowthBook service rather than waiting for that data to be made available in Athena

Hi there, we have been importing experiments through "setting -> data sources" but have noticed that our latest experiments aren't loaded even after we click the refresh list button. the "latest" experiment from the list shown has a start date of 24th Dec, while we definitely have newer experiments. Is this a known issue or potentially something we are doing wrong on our side?