Starbucks Internal Tools & AI Integrations (SaaS)

Written By Hao Trieu

***Page in progress, come back soon for more :)


 

Starbucks Loyalty AI-powered Campaign Builder

As a Senior Product Designer, I redesigned the entire Offers Campaign Builder product by myself (as a part of a larger tech migration effort), and turned extremely complex data elements, interaction models, logic, and an AI integration into an intuitive experience that our partners are having great success in using.

 
 

Context

Starbucks Loyalty is an essential and core facet of the Starbucks business, bringing in ~60% of its U.S. revenue from Starbucks Loyalty members.

Personalized Offers make up about ~50% of that Loyalty revenue, making it imperative for our Starbucks Marketing partners to be able to effectively create those campaigns.

We needed an all-in-one product that would enable the Loyalty business to continue thriving, and enable blue-sky flexibility to support any possible evolutions of that business.

 

Impact

  • Enabled the business to continue bringing in ~50% of its revenue, despite customer behavior and brand affinity changes

  • 100% adoption

    • Flexibility + guardrails and error handling help make the product as intuitive as possible

  • Unlocked the ability for Loyalty to gather data on new types of Offers that would be effective, and data for the AI model to provide better personalization through these Offers

 

Problem

Personalized Offers are EXTREMELY complex, consisting of many different systems, data elements, and configurations.

For any given Offer construct, there are around 10,000 possible variations, and the Loyalty team runs about 2-3 Offer campaigns a week. That’s about ~20,000-30,000 possible Offers that we can send to EACH of our ~35.5 million active Starbucks Rewards members, per week.

So how do we know which one to send, let alone create a campaign that provides value and incentive for our customers? How do we best optimize customer engagement and NIR - so that Loyalty can maintain its current revenue stream AND continue growing?

Starbucks did have a tool to create campaigns, but it was built on a back-end platform that was extremely limited. It didn’t allow us to create many types of Offer constructs, let alone effectively synthesize customer data and enable our AI model to provide effective personalization through those Offers.

Enter: Campaign Builder 2.0.

 

Opportunity & Goal

This Campaign Builder 2.0 initiative was started as a way to:

  • (1) unlock even the most basic and foundational Offer constructs that the previous tool couldn’t handle.

  • (2) redesign the experience to be more intuitive for our Loyalty partners to use and understand.

    • Make the front-end simple for something that is EXTREMELY complex on the back-end

    • Make the AI integration easy to understand, especially since it enabled different parts of the form depending on what mode it was in (fully optimized, hybrid, fully manual).

    • Loyalty Marketers engage with this product every day, so understandability of Offers throughout experience is important too.

  • (3) design it in a way that is as scalable and flexible as possible, for as many future evolutions of the Loyalty business as possible.

This also happened in parallel with a Loyalty redesign effort, so we needed to keep in close contact cross-functionally on what capabilities we needed to for sure prioritize, and how to prioritize them.

 

MVP Solution

***pictures coming soon***

“Offers creation” feature.

  • Offers creation

  • Offers summary

  • Offers edit

Dashboard (home for all created and draft campaigns).

Communication configurations.

Notifications configurations.

Audience segments.

 
 

Starbucks Next-Gen Point of Sale

As a Senior Product Designer at Starbucks, I’ve led and assisted in several areas for the Next-Gen Point of Sale system - including the creation of an uplifted Admin Functions dashboard and its first feature, strategy for how the POS will evolve, and laying down agile frameworks and holistic design documentation for our team.

 
 

Impact overview

  • Improved Design processes, with new documentation and documentation standards that allowed a new designer to onboard more quickly than before, and allowed our team to deliver features more quickly.

  • Created a strategy for an uplifted “Admin Functions” page, and delivered the experience for its first feature while strategizing how we will iterate into future iterations and be consistent with overall POS paradigms.

Hao Trieu
Previous

Starbucks Digital Gifting (B2C)
Next

Starbucks Loyalty (B2C)