Pattern’s Brand Managers can manage up to 10 brands, each with over 45 metrics to track. We found things often got lost in the shuffle because there was so much data and the Brand Manager didn’t know which metrics to focus on or what order to focus on them. Our goal was to build a tool so brand managers could quickly get an analysis on where each brand stood at a high level and then know where to drill down deeper.
With in Pattern many of our Brand Managers struggled to know where each brand was lacking. What we needed to find out was what metrics needed to be represented and which were the most important. We sat down with dozens of Brand Managers to help us identify what those metrics were and what ones were important to them. What we found was there was an order in how metrics should be resolved (An example of this is, if you were out of stock it didn't matter what your conversation rate was you needed to solve the out of stock issue before you worried about conversation rate) which was a big insight into how we needed to think about the metrics and how they were displayed.
We then identified what the order in which metrics needed to be solved and which high level metrics they were related to. This also lead to further understanding of what metrics we actually had data for and which ones we needed to dig deeper to find out more information about.
Understanding the user was key to this project, there were many stake holders for this project and making sure everyone was heard and their ideas were understood was important for this product to be successful. We found that for this organization that Maze user tests and In Person zoom calls were the best way to collect data from our users.
Collecting feedback through Maze was crucial to allow for fast iterations, I could send out a survey to 20 users and get responses with in an hour and have the feedback I needed in order to make updates to a design.
Periodically as the design would progress I would conduct user interviews to get more in depth feedback. These interviews allowed me to not only study what they would say but how they would say it was also a big indicator of what they were thinking and feeling about the product. This is where I was really able to see what they thought of the product.
Once a brand manager decided which brand to work on they then could use the Performance product to direct them where to spend their time. Performance would show them where the brand was excelling and where the brand needed help on a detailed and granular level. This empowered our brand managers to reach out to their clients and recommend actions for them to do in order to make all their Amazon listings improve.
‘Performance’ displayed data trends over time in an easy to read and follow format. Previously a brand manager would have had to dig through several reports to find this information. By making this historical data easy to access allowed brand managers to compare how brands are doing now vs a year ago. This also helped brand managers, and those over the brand managers, see if their efforts were helping a brand improve over time. Also it could be used to give insights across multiple brands so that a brand manager could see what was working with one brand and use that information to help another brand in their portfolio succeed.
Upon launch there was strong feedback that they wanted the feature to be more dynamic. We began the process of understanding how we could allow a user to simulate the numbers to be able to show what would be the monetary outcome for increasing specific metrics. What this would allow the user to do is be able to go to their brands and tell them this is what we need to focus on because it will have this much impact. This was still in the early stages of discovery when I left.
