Case Studies
Our Data Scientists Have Done Great Work For Great Clients
Here are some of the projects that our data scientists have worked on.
Prioritizing High Value Customers In Streamlining Call Centre Volume
Customer base of content company was segmented by value, recency of customers’ interaction, etc. As a result, company would then reframe all of their decisions from perspective of how these decisions would impact high value users.
Also, the company’s customer support call centre integrated the user segmentation to ensure that high value users’ questions are given priority in terms of wait time.
Also, the company’s customer support call centre integrated the user segmentation to ensure that high value users’ questions are given priority in terms of wait time.
Optimizing Customer Acquisition Advertising Budget
The client wanted to know how to optimize customer acquisition budget. At first, a revenue tracking system was built to track each acquired user cohort’s accumulated revenue, so that the client could focus on most profitable media sources.
Then, a complex regression model was built to predict the revenue that each cohort would achieve in a year’s time. This helped the acquisition team ensure that the advertising spend does not surpass the predicted revenue.
Then, a complex regression model was built to predict the revenue that each cohort would achieve in a year’s time. This helped the acquisition team ensure that the advertising spend does not surpass the predicted revenue.
Identifying Anomaly in Online Content
A promotional content publishing client wanted to make sure that if the content is scheduled to be posted that it is posted and is receiving traffic.
So, a system was built to ensure that for all the content that has been posted, there is a proper corresponding number of user views and clicks. If number of content views was too low, a flag would be issued to content team.
So, a system was built to ensure that for all the content that has been posted, there is a proper corresponding number of user views and clicks. If number of content views was too low, a flag would be issued to content team.
Identifying Problematic Parts of An App
The business wanted to know which parts of the app users struggled with and whether fixing those negative experiences would be valuable. So, survey data was combined with customer segmentation data.
As a result, the business could look at popularity of certain app features vs user satisfaction associated with those features. And, the business could then focus on improving parts of the app users use a lot, but with which they have a negative experience.
As a result, the business could look at popularity of certain app features vs user satisfaction associated with those features. And, the business could then focus on improving parts of the app users use a lot, but with which they have a negative experience.
See how we could help.
A no-obligation consultation