https://www.growthbook.io/ logo
Join Slack
Powered by
Do the Metric Overrides in an experiment behave li...
# announcements
m
Do the Metric Overrides in an experiment behave like the feature override rules? I’m hoping to run multiple experiments of the same feature with different populations(Franchise) as defined by my targeting attributes. For example: Test 1. Franchise attribute = 1 Serve 
Feature_A
with the values “Basketball” and “Baseball” Test 2. Franchise attribute = 2 Serve 
Feature_A
with the values “Baseball” and “Football”
f
hi, I’m not sure I understand exactly what you’re trying to do
is it one test?
or two tests?
m
haha, I’m not surprised. I’ve been struggling to figure out exactly what I need
f
heh
perhaps just a three way test?
m
so I work with a nation wide franchise. They want to allow each individual franchise the ability to test which tool works best for them.
I have that behind a feature in GB
f
interesting
m
I want to be able to run several tests concurrently for each franchise with different feature values
The feature overrides allow me to kind of do that, but then the test data would all go into one test
f
how many permutations are there?
yes, so one feature can have multiple experiments, and you can separate them if you need to - each experiment is tracked with the experiment_key
m
3 different permutations: v3 vs v4 v3 vs v5 v4 vs v5 But dozens of franchises who want to test
f
oh cool, that’s not so bad
saved groups might be a good way to opt the franchises into this
I’d be happy to chat quickly if you want to see how it could work
m
Yes, that would be great
28 Views
Open in Slack
PreviousNext
Powered by