Forum Discussion
See the history of changes to a custom attribute
Hello!
I would like to know if it's possible to see what, when and how often a custom attribute was updated. For example:
15/06
external_id: xyz
name: Jack
17/06
external_id: xyz
name: Peter
In that example I would like to see when external_id, xyz, changed name from Jack to Peter.
I'm asking as we have had a few cases where we believe users shouldn't have received an email from us and I want to verify when a custom attribute was updated and to what.
Hope that makes sense!
Cheers!
Hi jewsbur0TN - as far as I know, no… Braze does not provide the ability to review the history of changes for custom attributes on a specific user profile. Once a custom attribute is updated, the previous value is overwritten and there is no built-in mechanism to track or retrieve the historical changes of that attribute.
This is something you’d have to workaround with an external logging mechanism to capture and store attribute changes outside of Braze.
The only two things I can think of you could do within Braze directly are:
1) action triggered campaigns based on attribute changes - probably not helpful as you want to look back
2) webhooks capturing the attribute change and logging some kind of flag in the users’ profile - this will again not help you historically but if you can afford continue running your campaigns for a few days this could be useful to collect data on changes and re-evaluate, by segmenting customers on that flag. Of course this would imply you have very specific attributes to look at, as looking at all attributes would not only complexity this setup but consume a huge amount of data points
hope that helps!
Thanks inespais for the thorough answer. Would be annoying if there is no solution for this. I've also raised a ticket with Braze asking if they can provide a solution. I'll update here if they say anything.
Cool, would love to know if there's a way. They do sometimes have ways to enable certain features but have them disabled by default to avoid that your data point consumption goes wild, but have a feeling it won't be the case with this one 😐
