But this experiment is very important, so I just want to make sure I have this right.
We have 3 flags that we want to ensure get to the same users. If we use the same Bucketing Salt for each of the flags, then if we keep the segments and rollout numbers the same, then the same users should receive all 3 flags. Is this correct? Are there any pitfalls that we should be aware of when doing something like this?
Thanks in advance.
Best answer by Thao Nguyen
Hi Justin,
Thank you for your patience!
I consulted my team about your use case. You are correct in your understanding. If you use the same Bucketing Salt for each of the flags and keep the segments and rollout numbers the same, the same users should indeed receive all 3 flags.
This is due to Amplitude Experiment's consistent bucketing, which is based on the user, bucketing key, bucketing salt, allocation percentage, and variant weights. Given the same inputs, the output remains constant.
However, there are a few things to be aware of:
1. Changing the bucketing salt can cause users to switch between variants in your experiment. Therefore, it's recommended not to change the bucketing salt unless necessary. 2. If a user moves from one segment to another, they may be assigned to a different variant. This is known as 'Variant jumping'. To prevent this, you can enable a feature called 'Sticky bucketing', which ensures that a user will be assigned the same variant, even if they move between different segments. 3. Lastly, while using the same bucketing salt value for all experiments/flags will ensure that users receive the same variant, it's recommended to create a test experiment in a test project to observe the behavior before implementing it in production. I hope this helps! Let me know if you have any other questions.
Best regards, Thao
P.S. Checkout upcoming events and user meetups on our events page.
I consulted my team about your use case. You are correct in your understanding. If you use the same Bucketing Salt for each of the flags and keep the segments and rollout numbers the same, the same users should indeed receive all 3 flags.
This is due to Amplitude Experiment's consistent bucketing, which is based on the user, bucketing key, bucketing salt, allocation percentage, and variant weights. Given the same inputs, the output remains constant.
However, there are a few things to be aware of:
1. Changing the bucketing salt can cause users to switch between variants in your experiment. Therefore, it's recommended not to change the bucketing salt unless necessary. 2. If a user moves from one segment to another, they may be assigned to a different variant. This is known as 'Variant jumping'. To prevent this, you can enable a feature called 'Sticky bucketing', which ensures that a user will be assigned the same variant, even if they move between different segments. 3. Lastly, while using the same bucketing salt value for all experiments/flags will ensure that users receive the same variant, it's recommended to create a test experiment in a test project to observe the behavior before implementing it in production. I hope this helps! Let me know if you have any other questions.
Best regards, Thao
P.S. Checkout upcoming events and user meetups on our events page.
If you don't have an Amplitude account, you can create an Amplitude Starter account for free and enjoy direct access to the Community via SSO. Create an Amplitude account. You can also create a Guest account below!
If you're a current customer, select the domain you use to sign in with Amplitude.
If you don't have an Amplitude account, you can create an Amplitude Starter account for free and enjoy direct access to the Community via SSO. Create an Amplitude account. Want to sign up as a guest? Create a Community account.
If you're a current customer, select the domain you use to sign in with Amplitude.
We use 3 different kinds of cookies. You can choose which cookies you want to accept. We need basic cookies to make this site work, therefore these are the minimum you can select. Learn more about our cookies.
