We have a cloud instance of GB and we want to spin up a self-hosted proxy. Since it will be created in a

k8s

cluster, the other microservices will communicate with it internally. We can (additionally) expose its endpoint externally. My question is the following: What should be the value of

GROWTHBOOK_API_HOST

env var? I assume this should be set to the one that the GB cloud (managed instance) will try to reach it or the one that the other microservices will try to contact?

Hi everyone! I've got a question about ending experiments, and rolling out a variation to all users. Does one's selection in the 'Variation to Release' dropdown actually change how the feature is configured? The wording 'Which variation should be rolled out to 100% of users' implies to me that GB will update the feature such that all users receive either the Control or Treatment variation, but from what I can see this isn't happening. Did I misunderstand what that dropdown does, or is there an issue with having too many environments set up so GB doesn't know which to change?

If we have a cloud GB instance, where do we set these variables?

Hi, I’m looking for more info about using experiment phases. What is their intended use case? It seems like it is possible to add multiple phases to an experiment but what would that do to the data/analysis? All the documentation I’ve found suggests that it just filters out data before the beginning of the phase, so what happens if there are multiple phases? If users are assigned before the beginning of the phase but come back after it starts would all of that user’s data be filtered out, or only the pre-phase data? Can srm checks account for multiple phases with potentially different splits?

Is Slack integration gonna be free later for Open Source product?

which version of GB introduced (frequentist) multiple testing corrections? (also 🎉 !)

Hi Growthbook team! May I have a bit of advice? I wanted to start using experimentation in order to help make more data-informed decisions within my company. The company suffers from the classic “knee jerk” feature-build response to problems so, I wanted to introduce more rigour into what work is prioritised and subsequently built. Crappy ideas are cheap to demand, expensive to build! However, I believe the client’s sites just simply aren’t doing enough volume for quant methods to resolve and I have a stubborn “not enough data” lozenge within our test experiment results. Unbeknownst to me I had predominately been using “Frequentist” statistical approaches in the past at companies with a bit more scale in terms of ecommerce (Just Eat). I’m unfamiliar with Bayesian approaches (as in I haven’t hand-calc’ed any Bayesian approach to verify a set of results like I have with basic Chi-Squared or even simpler T-tests) but I just read your white-paper on the topic. So, my question; I got started with GrowthBook by doing my usual A/A test integration to ensure distribution and to correct our analysis SQL. What should I see if I were to pass the “not enough data” barrier (12 days remaining!). Is it a violin plot with a narrow distribution around 0%? The reason I ask is that I need to buy more time so that I can actually get onto the A/B without the whole idea being scrapped! More generally, what kind of volume does a site/app need in order to have these experiments resolve under 30 days. Is that even a fair question to ask of Bayesian engines?

Hi Team, We are using a Mixpanel integration with GB, and were wondering how we can pull in the "country" of a user, which is a property set by mixpanel. The

mixpanel.get_property

call only applies to super properties, and mixpanel doesn't seem to have a function to get non-superproperties into our SDK (and therefore Growthbook). Has anyone had any luck pulling this or other properties that are tagged at the event level? Thanks!

Hi Everyone, Could someone explain how I can add the ability to Stream Updates to a React connection. When hovering over the Streaming Updates: Disabled information icon it says to simply edit the connection, but when editing the connection nothing shows up for the Streaming of updates Thanks.

Hi Team, After creating growth book instance, i want to get all the assigned features of the user in the NODE JS code. could you let me know what approach i have to follow. If gb is my growthbook instance, gb.experiments is giving undefined value.

Hi everyone. We are starting to use GrowthBook and are wondering if there is an easy way to seed a containerised GrowthBook instance with features, users and sdk credentials. We’d rather not fiddle with the mongodb instance directly and wondered if there are any undocumented APIs, or otherwise recommended methods to achieve this, please?

Hello! My name is Elias, and I'm testing GrowthBook for Android. Following the documentation on the web, I see that the minimum Android SDK is 21, but when I implement it, it asks for SDK26.

This SDK supports both Java and Kotlin Android apps using Android SDK 21 and above.

Is this correct, or am I doing something wrong?

G'day! I'm evaluating GrowthBook (Cloud), integrated with a GA4/BigQuery data source. Question - is it possible to change the timezone used in reporting? It all appears to be showing UTC, ideally I would like to set this to

America/Los_Angeles

Hello, i'm testing the GrowthBook SDK for Javascript/React. We're encountering a weird issue with the

trackingCallback

that, for a user, isn't being called (it works for two other people). We've created a feature, enabled it, added an Experiment Rule which splits traffic 50/50, splits by id. What am I missing? Screenshots attached in this thread.

We have a case where we use a "Revenue" metric with denominator "All Experiment Users". In the same experiment, we use the same "Revenue" metric with a different denominator ("Orders") and we get different numbers on the numerator "Revenue". Do you know how can we further debug this problem? Any help would be much appreciated 🙌

Hi, I'm trying running the Growthbook on domain xxx.domain.com and api xxx-api.domain.com but all css and js return 404 error. Is someone command I need to build de css and js ?

Hi all, we had a question. We're self-hosting our version of Growthbook, but we are seeing 'This feature is a premium feature' tooltips. I saw on the docs here: https://www.growthbook.io/pricing that premium features should be available if we're self-hosting. How do we get access to the premium features when self-hosting?

In case of a self-hosted GrowthBook, what is the format the

MONGODB_URI

should have?

mongodb+srv://<username>:<password>hostname

?

if there is an env var named

PORT

, this is automatically used by both the

front-end

and

backend

and there is a conflict?

(in any case I assume there are 2 processes running in the

growthbook/growthbook

container, right?

I hava managed to set up my self-hosted instance on

k8s

. How to I see the ui to login? Because when I go to the url that exposes the services, e.g.

<https://my.growthbook.mydomain.com>

I see this

{
  "name": "GrowthBook API",
  "production": true,
  "api_host": "<http://localhost:3000>",
  "app_origin": "<https://localhost:3100>",
  "config_source": "db",
  "email_enabled": false,
  "build": {
    "sha": "cf216ccc76a9d54e4bd793781e3e1d2e45385548",
    "date": "2023-05-18T14:47:09Z"
  }
}

I am now able to reach the UI via `https://my.growthbook.mydomain.com`; What should I set the

API_HOST

/

APP_ORIGIN

env vars to?

From inside the container:

curl <http://localhost:3100>
{"name":"GrowthBook API","production":true,"api_host":"<http://localhost:3100>","app_origin":"<https://localhost:3000>","config_source":"db","email_enabled":false,"build":{"sha":"cf216ccc76a9d54e4bd793781e3e1d2e45385548","date":"2023-05-18T14:47:09Z"}}root@growthbook-api-web-685f59dbdf-fgxvn:/usr/local/src/app#

So why is the frontend unable to communicate with the API?

Is it possible to have to instances of the

growthbook

container running on different hosts, one just serving the

api

and the other one the

front-end

? (assuming we set the correct values for

API_HOST

and

APP_ORIGIN

?

Hi. I've learned that Growthbook could detect an already-used attribute ID so an experiment result will be the same for a particular user. I'm using growthbook in Golang backend and it works well in my local. My current implementation: • Keep a pointer of FeatureMap that will be updated by: ◦ A go routine, regularly fetched the latest feature values ◦ A webhook, register the webhook to Growthbook • On each HTTP request, initiate new growthbook context with attribute ID attached (userID). So, new growthbook instance for each request, using the same feature map. My question: • Where did this attribute ID being cached? How did the growthbook instance know that this or that ID is already assigned with an experiment variation? • If I spawn up 5 replicas of backend service, will the growthbook cached this attribute ID independently in each service? ◦ If so, will this lead to the same user receiving different growthbook experiment result in different service? Thank you in advance!

Hi Team, When setting up targeting conditions for the feature flags, is there a way to leverage Advanced Mode in order to have an attribute begin with a certain string rather than picking is equal to and other operators? Essentially

does match regex

Thanks.

I have set up self hosted api communicating with a self hosted proxy. Is there a way to validate this is OK before actually having an application making requests to the proxy? (no error logs in either the api or the proxy)

Hello GrowthBook users, I have a question regarding the client key. Is it safe or advisable to expose the growthbook client key on the client side? I would like to understand the potential problems or risks associated with exposing the client key in this manner. Your insights and experiences on this topic would be greatly appreciated. Thank you!

Perhaps this is a noob question since I am not very familiar with front-end. Is it "client rendered" in GB? This is why I am asking: Does the client try to reach the backend/api directly from the browser? (Setting the

API_HOST

to

<http://localhost:3100>

never worked despite the fact this url was working from inside the container)