Enrollment Blog

What Starbucks can teach enrollment leaders about using analytics to inform strategy

by Dana Strait

Walk down the street in any urban area, and it seems like there's a Starbucks on every corner.

I love coffee as much as the next person, but I often wonder how all of their stores can succeed. How do adjacent stores not cannibalize each other's sales? Don’t bespoke high-end cafes, local shops with loyal customers, and in-home single-brew choices (like K-cups) pose threats to the Starbucks model?

Turns out these are questions that nagged Starbucks executives earlier this decade. Their response provides surprising insight for enrollment leaders, particularly those looking to be more analytically savvy in their work.

A familiar problem: Volume growth but yield decline

From 2012 to 2016, Starbucks grew their total number of stores by 41%. But similar to some colleges seeing year-over-year application growth, the coffeehouse chain faced a significant "yield" challenge. In-store sales and average daily customers per store declined by 16% over this same period.

Starbucks sales

Fast-forward to today and Starbucks' same-store sales are back on the rise. We've compiled three lessons learned from their success to incorporate into your own enrollment management strategy.

Lesson #1: Digital innovation helps you be more responsive to key audiences

Starbucks was an early pioneer in using digital platforms to gather and respond to customer feedback. In 2008, they piloted MyRewardsIdea, a community website designed to crowd-source feedback and generate new ideas. Customers could submit or vote on ideas (for new products, in-store experiences, retail, etc.). Then when they implemented a customer concept, they let the community know, fueling further loyalty and engagement.

More famously, Starbucks first launched their mobile app in 2012. The app now boasts 17+ million users (including me!) and 90 million interactions per week. With an integrated order-and-pay system and a well designed customer loyalty and rewards program, they have scaled the kind of digital engagement they started with community groups.

By the end of 2017, more than 30% of Starbucks transactions originated from the mobile app, with app users spending approximately three times more than the average customer.

What 5,580 students shared about their digital communication preferences

The mobile app helped them reverse their "yield challenge," not just because it enabled more frequent customer engagement and simplified purchasing, but because it generated digital "breadcrumbs", behavioral data that their management team has used broadly to inform strategic decisions.

Lesson #2: Use digitally-generated "big data" to ask and answer critical questions

Because of the treasure trove of consumer data produced by the mobile app, Starbucks has changed the way they recruit and retain new customers. With 90 million customer interactions to analyze each week, they have real-time insight into nuances of customer behavior that were previously invisible.

For example, they now know how small shifts in weather will impact buying behavior. Surprisingly warm day in late winter? Amplify same-day promotions for iced coffee! They have data on geographical purchasing patterns and product preferences. They know what time of day coupons are most likely to increase customer engagement and for which segments.

At a more macro level, sophisticated customer modeling, or "lead scoring," helps them understand which customers to prioritize for promotions, what type of promotions will increase average customer spend, and who to target for new customer acquisition.

With the rise of Customer Relationship Management (CRM) platforms, enrollment leaders are sitting on a similarly powerful data asset that contains underleveraged (but powerful) information about prospective students. In my work as an advisor to enrollment leaders, I like to ask: what questions would you ask and answer of big data, if you were able?

Would you prioritize knowing conversion rates from prospect to inquiry? Inquiry to applicant? Would you want to know which students are most likely to persist through sophomore year? Which geographies you should target for search when your primary market becomes saturated? (The answer is usually yes to all of the above.)

Lesson #3: Accessing insights is hard. Applying insights is harder

But across all industries, getting those insights is easier said than done. Starbucks was not able to capitalize on their data overnight. Soon after the mobile app took off, they realized having the right organizational structure and talent would make or break their ability to leverage the newly available "big data."

They were quick to hire an ex-IBM analytics executive to run their first Business Intelligence unit to ensure strategy teams could nimbly apply insights against product, pricing, promotion, and positioning strategy.

Most colleges I talk to struggle to unlock the insights in their data for three reasons:

  1. The data required to understand prospective student behavior is often isolated, stored in inconsistent (and therefore incompatible) formats and lacking automated connections, which makes efficient analysis impossible.

  2. It is a competitive, expensive market for data science talent. Even with solid data, most colleges struggle to assign analysis out, especially with already overburdened institutional resource teams being the natural owner.

  3. Even for schools who do have accessible data at scale and the people to analyze it, bridging the gap from identifying important student insights to acting on them can prove difficult, without prescriptive insights and proven practices.

Introducing EAB's Pipeline Analytics platform

Given how frequently we heard these challenges and knowing that the successful application of "big data" requires resources most colleges don't have, EAB spent the last few years developing an enrollment analytics platform that gives enrollment managers real-time visibility into prospective student behavior.

Like WAZE (the traffic app) for enrollment, we want you to be able to anticipate enrollment risks and opportunities by having the data to truly understand prospect engagement, from first search through matriculation. And like Starbucks, we realize that analytics itself isn’t a silver bullet (knowing that first-generation students in Tennessee with an interest in nursing are best-fit for your school is useless if you can’t reach or engage with the right messages on the right channels), which is why it is integrated into our larger enrollment intelligence work, alongside smart recruitment marketing.

Even though selling coffee through a mobile app isn’t nearly as challenging (or important) as enrolling your next class, it’s always instructive to look at pioneers and innovators in other industries. The Starbucks story reminds us how a strategy that combines responsive customer engagement and analytics-driven decision making can lead to powerful results.

Here's how you can predict and improve yield

With the help of machine learning and pipeline analytics, universities can leverage behavioral data to improve enrollment outcomes. Learn how the latest data and analytics are shaping the future of evidence-based enrollment.

Download the White Paper


  • Manage Your Events
  • Saved webpages and searches
  • Manage your subscriptions
  • Update personal information
  • Invite a colleague