Question on using Canvas Flow for the Grubhub Team

To successfully collect customer sentiment, I utilized Braze Canvas in two ways:

1. In our net promoter survey, our goal is to measure customer sentiment from all kinds of merchants and highlight our highest value customers. To evenly collect survey responses from all types of merchants, we use Braze to segment by ‘tiers’. At Grubhub, our merchant partners are placed into different ‘tiers’ based on their performance on our platform. Ranging from tier A to F, our highest value customers make up the smallest percentage of our platform and have the lowest response rate. Our lower performing restaurants make up a higher percentage of our platform but have the highest response rate. In order to achieve the goal of collecting these net promoter scores, we wanted to get an even distribution of survey responses to ensure we are reaching all areas of the business. We worked with our MarOps team to create a custom attribute that recognized what tier a CID fell into and used that attribute to create an evergreen canvas. On a bi-weekly basis, our audience is entered into the canvas and funneled into 5 buckets (tier A,B,C, D, E&F). In order to extract an even amount of responses from each tier, each of them have to have a different % of a control group. For example, Tier A’s have the lowest response rate and also the least restaurants so it has a higher percentage of sends. Our Tier D’s have the highest response rate and the most restaurants so they have a lower percentage of sends. This allows for an even distribution of responses across the tiers. 
2. In our offboarding survey, our goal is to understand the reasons behind a merchant leaving our platform to improve our experience for existing customers. To achieve this, I worked with our MarOps team to trigger an email when a restaurant attrits from our platform. I was able to leverage existing attributes to identify these users and used them in our audience criteria.