Evaluation of Starbucks’ ‘Hello Again’ Campaign across Reach and the Customer Journey
TIMELINE
Fall 2025
ROLE
Group Member
Context
This was a semester-long market research project for COMM 3330 at UVA, completed as part of a three-person team. We set out to evaluate Starbucks' "Hello Again" campaign, a brand repositioning launched in February 2025 after several quarters of declining sales and a CEO change. I designed and ran the survey in Qualtrics, built the R analysis pipeline for both univariate and bivariate stages, and co-led the in-depth interviews and observational study. I approached the project like a real consulting engagement, structuring the research around a testable hypothesis that campaign awareness drives brand attitude, instead of just reporting descriptively.
Understanding the Problem
By mid-2024, Starbucks was clearly struggling. Comparable store sales dropped 6% and transactions fell 10%, leading to a CEO change that brought Brian Niccol from Chipotle. The company suspended its 2025 financial guidance after another poor quarter. In response, Starbucks hired Anomaly as its new agency of record and launched "Hello Again" in February 2025. This repositioning focused on three initiatives: personalized cup messages ("That's Not My Name"), encouraging longer in-store visits ("Stay Awhile"), and a free refill policy for brewed coffee and tea. coffee and tea.
Nine months into the campaign, there was still no clear evidence that customers had noticed these changes or that their feelings about the brand had shifted. Our main question was: how effective has Anomaly's "Hello Again" campaign been at changing how people see Starbucks?
Findings
Campaign awareness was very low: 71% of surveyed customers had never heard of "Hello Again" by name. However, among the 29% who had, we saw a strong positive effect on brand perception across every measured area: personal, authentic, and experiential, each at p < .001. Notably, customers who were aware of the campaign had an average NPS of 8.03, compared to 5.10 for those who were not aware. In summary, the campaign's reach was limited, but its impact on those it reached was strong and measurable.
Awareness was not the only factor. The best predictor of whether someone would recommend Starbucks was how personalized their experience felt (Spearman's ρ = 0.62, p < .001). Small touches, like handwritten cup messages, did more for brand loyalty than the broader "Stay Awhile" idea. Our exploratory research confirmed this: when people finished the sentence "A personalized message on my cup makes me feel...," their answers were overwhelmingly positive: happy, seen, special, valued, supported.
The data showed a clear tension in the campaign. "Hello Again" aims to get people to stay longer, but 57% of customers spend less than 10 minutes in the store. The groups most engaged with "Stay Awhile"—workers who stay over an hour and socializers who stay about 24 minutes—are real but small. The factors that most improved NPS and personalization ratings were the pickup and app experiences, which matter most to the grab-and-go majority. Right now, Starbucks is not focusing on the top consumer priorities from our survey: fast service and high-quality products.
We cannot completely rule out that people who already like the brand are more likely to notice the campaign, rather than the campaign creating new brand loyalty. However, the size of the differences—almost 3 full NPS points and a full point on the experience scale—suggests the relationship is not just due to selection bias.
Methodology
We carried out four phases of primary research, with each phase building on the last. We started with informational interviews to build vocabulary and spot themes. Next, we conducted 11 in-depth interviews with customers aged 13 to 81, using a guide that progressed from specific questions about products and ordering to broader questions about functionality, psychology, and values. We chose in-depth interviews over focus groups to avoid peer influence in how people described their routines and habits.
At the same time, we observed 57 customers at a Charlottesville Starbucks on a Sunday afternoon, recording their age group, order type, time spent in the store, and main behavior. This led to three behavioral segments—workers, socializers, and to-go customers—which shaped our survey design. We also used a projective technique to understand emotional responses to the "That's Not My Name" initiative.
The final phase was a 12-question structured survey given to 101 U.S. respondents through Amazon Mechanical Turk. The survey was designed with a clear independent variable and dependent variable structure: we measured attitudes toward Starbucks (personalization agreement, three semantic differential scales, NPS) first, then asked about awareness of campaign initiatives, using a funnel method to avoid priming. We analyzed the data in R, starting with univariate summaries to set baselines, then using t-tests, chi-square tests, one-way ANOVA, and Spearman correlations to see if awareness predicted perception. This approach allowed us to move from general impressions to specific, testable claims about which initiatives were effective.
Thank you to Ella Butensky & Piper Toney for working with me on this project!