Hello Everyone! Happy New Year! 🍾 We have just released version 2.7. Here are some of the new features we've added. • Sticky Bucketing for JS SDK users. Sticky Bucketing allows you to remember the first variation a user sees, so their experience will remain consistent, even if something changes that would otherwise re-assign them. This is an opt-in feature. (Read the docs) • Reusable Targeting Conditions. Just like the saved groups, you can now save targeting conditions to make targeting new features to previously targeted users easier. (docs) • Huge improvements to the fact table queries. For our enterprise customers, we've greatly improved the efficiency of queries when using fact tables. (blog post about it) • All experiment reports now have a health tab. You can see SRM warnings and other information about the health of your experiment. (docs) • Safely Update Live Experiments. Changing a live experiment mid-flight is much more complicated than many people realize. This feature guides you some of the issues and helps you make the right decision for your experiment. Read the blog post about this feature • New Best Practices Guide. We've added a new section to our documentation site which contains guides on experimentation in general and chapters specifically on how to get the most out of GrowthBook. Link • Contextual AI bot for helping you answer your questions, trained on all our documentation and systems. And lots of other bug fixes and improvements, Okta SCIM improvements, and more. You can read about it on our blog post: https://medium.com/growth-book/new-growthbook-version-2-7-ea4d0e8c2ed6 or on our GitHub: https://github.com/growthbook/growthbook/releases

Hello, I'm setuping growthbook as self hosted version on kubernetes. I create one deployment for ui and api with only 1 replica and an other deployment for proxy with 3 replica (for a stress test) I have few questions • api use to crash. ui app work but without api I just have "fail to fetch", here the logs

yarn run v1.22.19
$ wsrun -p 'back-end' -p 'front-end' --no-prefix -c start
$ next start
$ node dist/server.js
ready - started server on 0.0.0.0:3000, url: <http://localhost:3000>
{"level":30,"time":1702482550777,"pid":60,"hostname":"deployment-ui-9c56d45c7-wkhl6","msg":"No config.yml file. Using MongoDB instead to store data sources, metrics, and dimensions."}
(node:60) [MONGOOSE] DeprecationWarning: Mongoose: the `strictQuery` option will be switched back to `false` by default in Mongoose 7. Use `mongoose.set('strictQuery', false);` if you want to prepare for this change. Or use `mongoose.set('strictQuery', true);` to suppress this warning.
(Use `node --trace-deprecation ...` to show where the warning was created)
{"level":30,"time":1702482551567,"pid":60,"hostname":"deployment-ui-9c56d45c7-wkhl6","msg":"Back-end is running at <http://localhost:3100> in development mode. Press CTRL-C to stop"}
{"level":30,"time":1702483922168,"pid":60,"hostname":"deployment-ui-9c56d45c7-wkhl6","msg":"Start Refreshing Results for experiment exp_8kpn21nlq3kyzxc"}
Killed
error Command failed with exit code 137.
info Visit <https://yarnpkg.com/en/docs/cli/run> for documentation about this command.
`yarn start` failed with exit code 137

• I saw in doc to set PROXY_HOST_PUBLIC=https://growthbook-proxy.example.com I think it's for invalid cach, if we have multi proxy clusters (1 by kub cluster) can I add multilple URL ? • in proxy log I have this : "msg":"connection polling: no connections found" what is it about ? thanks Geoffrey

Hi there! 👋 Can someone please point me in the direction of docs for general advice on mocking feature states in a React app? I want to my components for when a flag is on or off, for instance.

Hey guys I have a question about self hosting. The current requirement to have separate ports is a bit confusing to me. Is it not possible to host both over 443? I currently have the front end on 443 and the backend over 80 but this is throwing a lot of security concerns.

Hi everyone. Have a question regarding Growthbook installation via GTM and event tracking with GA4 (which is also installed via gtm). Initialization code installed with the SDK (I`ve tried to do it separately with the same results) CSP allows the launch of the script. AA test is launched in a growthbook control panel. The main problem is I can`t send the ga4 event on callback. Secondary problem - I can`t get any info from growthbook debugger (dev mode is on) because it isnt working. What do you think I should do to fix this issue? Has someone already faced this problem?

I'm using growthbook with a nextjs app... Weirdness is going on where locally I get the experiment but on production I don't...

const growthbook = new GrowthBook({
  apiHost: process.env.NEXT_PUBLIC_GROWTHBOOK_API_HOST,
  clientKey: process.env.NEXT_PUBLIC_GROWTHBOOK_CLIENT_KEY,
  enableDevMode: process.env.NODE_ENV === 'development',
  trackingCallback: (experiment, result) => {
    console.log('Viewed Experiment', experiment, result);

    if ('gtag' in window) {
      (window as any).gtag('event', 'experiment_viewed', {
        event_category: 'experiment',
        experiment_id: experiment.key,
        variation_id: result.variationId,
      });
    } else {
      console.warn('no gtag');
    }
  },
});

locally i get the console log for Viewed Experiment. On Prod I don't get any logs at all... Very strange... The only difference should be enableDevMode (I tested locally the production build...)

I am setting up GrowthBook Proxy and the documentation does not seem clear on various points. It mentions

SECRET_API_KEY

and to setup in Settings -> API Keys. Does that need to be an

Admin

key or can it be just a

Read-Only

key?

Hey team, just wondering if Growth Book offers multivariate testing functionality also ??? Could not get the answer for this in the support document.

Hi everyone, I am setting up "identifier join tables" and I have a doubt. • let's say that I am randomizing on

client_id

(cookie) • I have a

lead_id

that is generated once some step is done • the

lead_id

is the key used to identify conversions • so i set up a join table like

SELECT client_id, lead_id FROM table_relation

here are the doubts: • Is there a smart way that I am missing to handle repetition? example: 1. I am on device 1 (so I have

client_id_1

) and generate a

lead_id_x

. 2. then I log in on device 2 (so

client_id_2

) retrieving

lead_id_x

so same lead_id from point 1 3. then i do a binomial conversion with my

lead_id_x

4. when doing analysis, I would have for both

client_id_1

and

client_id_2

a conversion (because are both linked to

lead_id_x

• Is there a way to label these phenomena in analysis (ideally, I would also like to count distinct conversion)? • Do I have to create some attribution logic to handle it? e.g. only last

client_id_n

if multiple...

Hey I want to set forced value for my feature using the sdk, how can I do it? Thanks. I found this ref in the docs: https://docs.growthbook.io/lib/js Still not clear about the implementation.