7 lessons for developing a customer personalization engine 

“This blog post highlights 7 lessons our team learned as we developed and deployed a data-driven personalization engine.”
Laila Haerian ImageVineeth Pillai Image
By 
Laila Haerian & Vineeth Pillai
November 10, 2023

The digital analytics team at Discover® Financial Services recently created a tool that used data, statistical models, and algorithms to recommend better marketing offers to customers.

This blog post highlights seven lessons our team learned as we developed and deployed the project.

1. Start small

The personalization product started with a single test that ran in one of the most prominent regions on the Discover’s Mobile App. That small test used a simple, straightforward algorithm and a scaled-back platform, but it indicated that the idea would be successful on a larger scale. 

2. Give your team ownership

Our team has a profound sense of ownership over the results and is committed to its success. The analysts who built the engine have been with this program from day one, and—because they have ownership and leadership support—have been able to innovate and grow the program with great results. 

3. Choose the right tools 

After evaluating numerous platforms and tools that were widely used in the industry, there was not one that met all the project’s needs and requirements. Our team chose tools that gave the most benefit and room to grow, and then customized the tools and rewired the platforms to fit Discover’s specific standards and ways of working.

Our team needed to implement the necessary rigor around decisioning, robust test-and-learn agenda, and automation — and all behind the Discover firewall.  

4. Have a deliberate growth path  

The analysts mapped out a deliberate path to growing this program across different regions and different channels on the Discover site and app. So, as the team onboarded different partners, it was also onboarding different regions, expanding the program’s scope systematically. 

5. Measure effectiveness and incremental value

We have a solid analytical framework in place to measure the effectiveness and the incremental value that this program generates. Strategy changes are made based on data insights, and every change is tested and evaluated for its effectiveness. The result is an ever-evolving product that serves customers no matter what is happening in the broader economy.   

6. Evangelize through quality results

The program grew organically because the team maintained the level of quality that it promised. Other teams at Discover were more willing to take part, seeing the quality results. The emphasis on customer experience, quality of deliverables and transparency of the process helped rally the cross-functional teams to roll out the functionality. 

7. Put yourself in your customers' shoes

Even with emphasis on offer recommendation, focus on customer experience was always a top priority. While the offers that are shown to customers are financial products and services, there are personalized messages reminding cardmembers to activate their card, alerts about security features they should consider, or even prompts to encourage selecting a new card design. These small things to delight customers were a key aspect of this approach, ultimately helping them to enjoy their experience with Discover. 

Conclusion

Discover is committed to creating bright financial futures for its customers. The digital analytics team’s personalization product ensures that customers receive the most pertinent offers at the exact right moment in their financial journeys, which is imperative to building a stable financial foundation.  

© 2023 Discover Financial Services. Opinions are those of the individual author. Unless noted otherwise in this post, Discover is not affiliated with, nor endorsed by, any of the companies mentioned. All trademarks and other intellectual property used or displayed are property of their respective owners