Forum Discussion
Segment Builder funnel stats design
Hi everyone! I'm Nancy, the designer at Braze working on our segmentation tools. We've received feedback about wanting to see funnel stats in our updated segment builder design to easily track reachable user increases and decreases. As I work on this design, I'd love to get your feedback, and I have three questions that will help inform our design decisions.
Each question is associated with the screenshot directly below it, and we recommend you answer the first one before viewing the second screenshot, etc (because later screenshots will "give away" some info). Thank you in advance, and let me know if you have any questions!
Question 1
Looking at this image below, could you tell us what you would do if you wanted to see per-filter funnel stats (ie. what step you take, where would you click to view this info)?
Question 2
Looking at this image below, the second filter group contains two filters, Language and Age. What would you expect the number "350k β 400k" under the Language filter represents?
- 350k - 400k is the number of all users that fall under the specific Language filter, irregardless of whether they fell into prior filters (in other words, does not take into account prior filters)
- 350k - 400k is the number of all users that fall under the Language filter and the prior filters in filter group 1 (in other words, it takes into account all filters up until that point)
Question 3
Looking at this image below, what would you expect the red number in the group footer to represent:
- It represents the estimated user that fall under all the filters from group 1 and 2
- It represents the estimated user in group 2 only
- ArsoExpert
- You have a lot of real estate in the filter rows. We all want a clean design, but those numbers are something most of us Braze users found valuable in the old UI, so there would be value in showing stats in each row and not having the design be as "clean".
Ideally, an extra step wouldn't be required to get the info that was once visible by default. So, no clicking. In the name of clean design, I wouldn't be opposed to having an option to show/hide these numbers on the segment level. But segment-level only, not for each filter. - Nicely done! To answer the question - based on how you've set up the segment, only the first option makes sense. You have an OR operator between the two filter groups, so the first filter group has zero impact on the audience size defined by the second filter group. I think those should always be independent of prior filters. The only scenario where it would make sense for it to be the second suggestion, is if you use only AND operators in both filter groups and you have an AND operator between the groups - in which case you wouldn't need to use groups π
I have some feedback on the numbers as well. In the past, this used to be a single number, not a range. While a range might be a more accurate representation than the single-number estimations of the past, some of the ranges I see on the screenshots are not what I'd consider helpful. 350k-400k is a lot of variance. Same for 420k-528k in the second filter. I hope the ranges are random and only for the purposes of the demo. - Group two only.
I have some more feedback on the labels and where to show the estimations. I think that you should either have the group estimation where the red number is or below the group, but not in both places. The estimation placed next to the OR operator between the groups looks a bit out of place.
I just realized that you can hide the group estimations and that the ones after (and outside of) each group show the cumulative total. I still feel like they're a bit out of place, especially when having the independent group estimations shown + the "per channel" stats for each group. It's cluttered.
The label "estimated reachable users" is misleading because that number has always been a sum of reachable and unreachable users. Even if you take the estimations at the bottom - the first column "Total" is always showing reachable + unreachable users. Then, each channel column shows the users that are reachable on that particular channel. In the example shown in the screenshot, the "Total" and "Email" columns would have the same range (or exact stat) only if you use an "Email available is true" filter. "Estimated users" would be more accurate.
I'm not sure if showing the number of reachable users per channel on each filter group is helpful, but I don't mind having it there.
- nancychiuBraze Employee
Hey Arso! Thanks so much for your feedback! Seeing your detailed thoughts has been really helpful!
Regarding your response to question 2 - yea I messed up in the design mockup π It should've been an AND between the groups so that it'd make sense for the funnel stats. And yes, the range in reality will be much smaller than what I have in the mockup, we're exploring range to further distinguish estimated numbers from exact numbers.
Regarding your comment on "estimated reachable users", I double checked with our engineers, and confirmed that the total number is correctly reflecting total reachable users (and no unreachable users). The reason why sometimes the total number and each channel breakdown number is different is because there are additional channels such as SMS that we're not showing in the channel breakdown.
Thanks again for your feedback!
- You have a lot of real estate in the filter rows. We all want a clean design, but those numbers are something most of us Braze users found valuable in the old UI, so there would be value in showing stats in each row and not having the design be as "clean".
- JFCRobinSupporter
Question 1
Looking at this image below, could you tell us what you would do if you wanted to see per-filter funnel stats (ie. what step you take, where would you click to view this info)?I'd guess at either clicking the bar chart icon, or just hovering over the 'estimated reachable users' text and hoping my mouse flips to a hand and I could click / expand it.
Question 2
Looking at this image below, the second filter group contains two filters, Language and Age. What would you expect the number "350k β 400k" under the Language filter represents?I'd expect it to be the estimate of only those contacts matching English OR Japanese NOT the cumulative estimate of all filters so far.
Question 3
Looking at this image below, what would you expect the red number in the group footer to represent:I'd expect it to represent the estimated users from group 2 only
- nancychiuBraze Employee
Thanks so much for your feedback!
Related Content
- 3 months ago
- 2 months ago