facebook pixel

Bucketing - Feedback categorization

Bucketing Ask Inline's way of automating customer feedback categorization. Each bucket has a name and a set of filters. Filters determine what feedback items are placed into a bucket. Feedback items can belong to multiple buckets and you may create as many buckets as necessary.

Why categorize feedback?

After a baseline set of feedback is collected, we recommend reading through several dozen items in order to develop a set of categories that the feedback falls into. Creating categories allows teams to track how customer feel about specific aspects of their application or service are changing over time and how the sentiment of distinct customer segments develop over time.

Creating a new bucket

Buckets are administered and viewed though the buckets tab inside the Ask Inline data dashboard.

While logged into the the Ask Inline data dashbaord, navigate to the "Bucketing" tab in the header.

Buckets have a name and a set of filters. Provide a name for your new bucket and click the Add filter button.

Image points out the location of the Bucketing tab in the Ask Inline data dashboard.
The bucket name field and Add filter button in the data dashboard.

For the Bucket's fiter, provide a list of keywords that you would like to associate with this bucket. This bucket will only contain feedback items which match at least one of these keywords. After the keyword list has been provided, click the save button.

Providing a comma separated list of words for the Bucket to filter by and saving the newly created bucket.

Understanding the bucket display

Buckets track feedback items that match it's filters independent of the rest of the system. This allows buckets to create an NPS or CSAT trendline that is specific to the feedback that matches those filters. This effectively allows for category based scores. For example a product team can track not just the overall NPS, but the design and performance NPS of their product by creating buckets for those two categories of feedback. While this isn't a "true" NPS, it is useful for seeing trends: "All of our detractors complain performance problems. A performance sprint will probably help our overall NPS score".

A performance bucket and it's respective NPS trendline. Unlike the homepage of the data dashboard, which show a ten week rolling average, buckets only display a seven week rolling average of it's scores.

Example performance bucket in the Ask Inline data dashbaord.

In addition to the NPS trendline, buckets also display the number of items form each week that fall within the bucket. This allows teams to understand the volume and frequency changes to a bucket. This is useful when for understanding how work is impacting a specific category. For example, imagine a company changes the way it's support phone que is processed in an effort to ensure customers recieve help more quickly. Before introducing this change, they might create a "Wait time" bucket to their Support Call CSAT campaign. Hopefully, after the change is put into place, the number of customers mentioning long wait times will drop.

A wait time bucket and it's respective volume indicators. These show trends over the past seven weeks of data for this bucket.

Example wait time bucket and it's volume indicators.