Cricket Wireless

Deep learning drives Cricket device sales and 7X in-store visits

Overview Client
Industry Wireless Phone
Program Application Mobile Display
Product Cognitiv Incremental
Key Business Challenge

Drive in-store visits through online interactions

Identify new customer segments

Improve measurement of ad effectiveness

Client objectives

Drive in-store visitation of non-Cricket customers through a mobile display campaign during the 2017 holiday season.

High-level Results
0 %

Cut Cost Per Incremental Visit

0 x

Drove Incremental Lift on In-Store Visits

0 %

Improved Targeted Incremental Visits

0 %

Non-Incremental Visits Cut Costs

The Story
The Story

Seasonal sales soar with custom deep learning.

Implemented Cognitiv Incremental during the 2017 holiday season to ensure holiday advertising identified new customer segments and untapped profits to drive in-store visits.

Engaged first-party, mobile user data combined with Cognitiv’s core technology, NeuralMind, to automatically train a custom deep learning algorithm tailored to discover non-cricket devices, highly likely to visit Cricket’s stores.

Cognitiv’s custom deep learning algorithm drove a 7X Incremental Lift on In-Store Visitation for Cricket and decreased the cost per visit by over 50%.

Solutions
  • Develop a training data set of all mobile devices that visited Cricket stores in the past 60 days.
  • Create a custom deep learning algorithm tailored to discover non-Cricket devices highly likely to visit Cricket’s stores.
  • Execute a custom deep learning bidding algorithm to automatically bid and optimize on programmatic exchanges and advertising performance in real-time.
Performance
  • Drove the most successful lift campaign in the history of the retailer, with 773% increase in incremental in-store visitation.
  • Execute dynamic bids on every impression to deliver a reduction in CPV by 60% and Incremental CPV by 50%.
More Case studies
Discover what the brain behind the brands can do for yours. Learn more
Cognitiv
Lorem ipsum
Lorem ipsum