Effects of proposed changes to address unattributed trials at Treehouse

This document helps to understand marketing conversion and how to address unattributed trials as part of our marketing processes. It includes recommendations and actions we took to combat the issue of non-attribution.

Alexis Johnson-Gresham
Lead Product Data Analyst at Reforge Previously Brooklyn Data Co., Treehouse, Precision Strategies, PwC

One of the things I worked on as a product analyst was helping to understand marketing conversion.Attribution is a really important part of that, but we'd seen large increase in unattributed trials. This analysis and proposal was to uncover whether there was a technical issue, to propose process and tech changes, as well as to show the impact of those propod changes.

Treehouse_SLIDE 1

This slide is so important!

There is a lot of information in this presentation, and the presentation served as a piece of documentation for marketing and data. I wanted to make sure people got the highlights before we got too deep into any detail, as well as ensure that the context would be clear when we looked back.

Even though I had clear suggestions for what we'd do overall, there were additional items that marketing and data needed to resolve together.

Treehouse_SLIDE 3

This and the next few slides were all about showing the scale of the proposed changes.

Rather than just telling the team that the changes were important, the visual references helped magnify why we should prioritize the time needed to make the changes and made sure that the decision makers had a clear understanding of critical nuances.

Treehouse_SLIDE 4
Treehouse_SLIDE 5
Treehouse_SLIDE 6

Treehouse_SLIDE 7
Treehouse_SLIDE 8
Treehouse_SLIDE 9
Treehouse_SLIDE 10

We all agreed that this was a problem—this slide shows our lingering questions that we needed to answer to make the fix.

Treehouse_SLIDE 11

The to-do list for data teams is often long. Connecting the dots between previously discussed needs and current needs was helpful for our team for prioritizing.

Data is never just about things the data team is doing or owns—it's important to connect the dots between upstream elements that other teams own. This one represented something that would need to be worked on across engineering, data, and marketing.

Treehouse_SLIDE 12

Sometimes you can do anything about missing data, and sometimes you can't. In this case, it was important for the team to know that we couldn't do anything about this particular segment of data.

Treehouse_SLIDE 13
Treehouse_SLIDE 14
Treehouse_SLIDE 15
Treehouse_SLIDE 16
Treehouse_SLIDE 18
Treehouse_SLIDE 17

Add reaction

Add reaction

Add reaction

Add reaction

Add reaction

Add reaction

Related Artifacts