Brewing Festive Flavors: A Digital Dive into Renaming the Peppermint White Chocolate Mocha

 

This project focuses on designing an experiment as a Starbucks Marketer. The scenario involves whether a drink name change would boost sales for the holiday beverage. The question being focused upon is whether this name change will increase revenue for the brand, specifically pertaining to that holiday drink. In the prompt, we are tasked with the idea of changing the “Peppermint White Chocolate Mocha” beverage to “Winter Wonderland Mocha”. In this assignment, I will focus on how I will conduct the experiment when it comes to recruitment, experiment design, and measurement of data, the control and treatment conditions within the research design, the “why” behind choosing between-subject design, and confounding variables and how to minimize them. Each focal point has been broken up into digestible sections for clarity and organizational purposes. After exploring each of these areas, an infographic I created using the Canva platform is provided, highlighting the scientific research method being used.

Tap the infographic below to explore a summary of the project using the scientific method: 

How the experiment will be conducted:

Recruitment:

To start an experiment, it is first necessary to recruit participants. Starbucks has a large number of users on its app, in fact their “mobile payment app is so popular that a few years ago it had the most mobile payment users in the United States” (McKinnon, 2023). Considering this, leaning into the app to recruit participants for the experiment would be a great opportunity to draw from a large group of individuals. This can be done simply by having a pop up when one opens the app, or notification prompting one to participate. The notice would utilize random selection to participate in the experiment to determine which drink name is stronger and more marketable leading to higher conversion rate, the point at which an individual purchases the product. Although, the participants will not have the awareness that this is the question prompting the research and experiment. Instead, users will be introduced to a fun holiday drink game where they will earn rewards for participating. Incentivizing engagement will be crucial in the recruitment process as it has “a crucial role in market research by motivating study participation” (Trappe, 2024). For this reason leaning into the Starbucks Rewards system will be beneficial to recruiting participants for the experiment. Finally, size will be important when examining credibility of the experiment being conducted. Aiming for a large size of participants, for example, 10,000 - 20,000, should not be too challenging considering “more than 31 million people use the Starbucks app” (Aten, 2024). 

Experiment Design:

Random Assignment will be necessary to incorporate as an “important part of control in experimental research, because it helps strengthen the internal validity of an experiment and avoid biases” (Bhandari, 2023). In my opinion, the use of gamification, “use of game design elements in non-game contexts” can serve as a strong incentive to drive user participation (Khoshnoodifar, Ashouri, & Taheri, 2023). Drawing from my own experience with the Starbucks app, I really enjoy the games that they incorporate into the app. Examples include the Starbucks for life game where one needs to complete certain tasks for a chance to win Starbucks for life or earn points for free drinks or food. The pairing of gamification for the purpose of this hypothetical research scenario is great as it can boost user participation in the experiment. 
Going back to the mention of random assignment as a crucial piece of the research, when leveraging the app, an algorithm can be used to randomly assign participants to either the control group or the treatment group. A fun way to incorporate a gamified element to this process would be to unwrap a present, similar to previous Starbucks games, or shake a snow globe, as done previously in the winter version of Starbucks for Life. This continues alignment with holiday eir of research (after all, one is drawing data about the names between a holiday beverage) while also providing an interactive element. An example is provided below: 

Image Description: Starbucks for Life game. Snow-globe with coffee in center and holiday decor
Reference for above image: https://pin.it/4375dS4GF

Control and Treatment Conditions:

The experiment will have a control group and a treatment group. Both groups will incorporate gamification aspects. Within the control group, participants will see the drink labeled as “Peppermint White Chocolate Mocha”. Here, control group members will see their points and earnings with the Starbucks Holiday game theme for the experiment. On the other hand, writing the treatment group, users will see their point drink listed as “Winter Wonderland Mocha”. This drink will be listed under the same holiday theme visually and the participants within the treatment group will still be able to earn points for the “Winter Wonderland Mocha” beverage. Earning points or gifting points dependent on the way the game is going for each participant will be the same across both groups so as not to incentivize buying one drink over the other. For example, if a participant in the control group is awarded 30 points for getting a “peppermint white chocolate mocha”, a participant in the treatment group will be awarded the same amount of points to get a “Winter Wonderland Mocha”.  

Measurement:

There are several forms of measurement that can be incorporated in this experiment. First, tracking sales or purchase of products will be essential to determining the conversion metrics between the “Winter Wonderland Mocha” compared to the “Peppermint White Chocolate Mocha”. Since an app is being leveraged to conduct the research, the analytics from the app can be used to monitor these metrics. Another way to measure could be with a survey. Although surveys can seem monotonous and bland, if incorporating gamification, it could be a fun way to draw more information for participants. For example, once one purchases the “Winter Wonderland Mocha” or “Peppermint White Chocolate Mocha” there can be a pop up survey that visually sticks with the holiday theme prompting users to identify sentiment associated with the drink name in exchange for more points added to their app. This would drive up user feedback, as it has shown to assist “companies better understand what people need and expect from a product or service” (Krysiak-Adamczyk, 2024). The juxtaposition of sales analysis and surveying for user feedback generates solid data for evaluation of Starbucks Holiday beverage perception. 

Research Design: Standard A/B Testing Paired with Gamification

The research design will use standard A/B testing paired with gamification. Standard A/B testing is a strong design method to lean into as it can be used to “determine the improvements in the user experience that are best for their business goals” (Neusesser, 2024). In the case of this experiment, it will be used to determine which name is a stronger choice for the holiday beverage- “Peppermint White Chocolate Mocha” (control condition) and “Winter Wonderland Mocha” (treatment condition) both with the same gamified aspects and and holiday visual design appeal. 

Between-Subject Design Choice:

Between subject design will pair well with the research design as each participant is only exposed to one condition. In other words, this research design requires dividing “the sample into two groups” (Simkus, 2024) which is precisely what is being done in this experiment, as the control group is being exposed to the term “Peppermint White Chocolate Mocha” while the treatment (experimental) group is being offered the term “Winter Wonderland Mocha”. 

Potential Confounding Variables and How to Minimize them:
Some confounding variables include the seasonality of the beverage. The drink name that is being researched is only available as a part of the holiday/winter season. This impacts the sales variation. A way to minimize the implications of this includes conducting the experiment during the seasonal time so that it goes along seamlessly with the holiday drink offerings; this controls the seasonality confounding variable. In addition, marketing of the seasonal beverages in general may act as a confounding variable. If you are familiar with Starbucks seasonal offerings, when they drop their holiday menu it’s a big deal, not to mention the free reusable Christmas cup day (Starbucks, n.d.). The prompting of consumers to purchase holiday beverages, especially the one being tested (Peppermint White Chocolate Mocha) could have impacts on the experiment. One way to minimize this may be that general marketing does not name drop the Peppermint White Chocolate Mocha and that if notifications are issued via the app prompting a consumer to buy a holiday beverage, its associate with the same name in their group (i.e. control group will receive the original name, treatment/experimental group will receive the “Winter Wonderland Mocha” name). 
To add, app usage varies across Starbucks consumers, therefore this has the potential to act as a confounding variable. For this reason, it would be essential to consider user engagement through app analytics when reviewing the data or recruiting participants to minimize this impact. Finally, because gamification is leveraged across this research/experiment design, it is important to consider this as a confounding variable- some users may be influenced by the gaming aspect with regard to their purchasing pattern. This confounding variable that is brought on by gamification is minimized by using the same reward system across both groups, which is discussed earlier when laying out the research design. Each of the confounding factors should be considered and opportunities to minimize them should be implemented so that the integrity of the testing can be maintained. This will allow for cleaner data to be extracted so that the question of which holiday beverage name should be used for optimal user experience and consumer conversion can be answered and appropriate changes to the name can be made! 

References:
Aten, J. (2024, February 20). Starbucks made a simple but brilliant change to its app ... Inc. https://www.inc.com/jason-aten/starbucks-made-a-simple-but-brilliant-change-to-its-app-that-every-customer-can-appreciate.html
Bhandari, P. (2023, June 22). Random assignment in Experiments: Introduction & examples. Scribbr. https://www.scribbr.com/methodology/random-assignment/
Khoshnoodifar, M., Ashouri, A., & Taheri, M. (2023). Effectiveness of Gamification in Enhancing Learning and Attitudes: A Study of Statistics Education for Health School Students. Journal of advances in medical education & professionalism, 11(4), 230–239. https://doi.org/10.30476/JAMP.2023.98953.1817
Krysiak-Adamczyk, A. (2024, November 29). Complete guide to user feedback (types, how to collect+analyze). Survicate. https://survicate.com/blog/user-feedback/#:~:text=iterative%20product%20improvements.-,Why%20is%20user%20feedback%20important?,perspective—that%20of%20your%20users.&text=In%20our%20recent%20report%2C%20Impact,customer%2Dcentric%20designs%20and%20features.&text=By%20gathering%20user%20feedback%2C%20companies,product%20and%20resolving%20usability%20issues.&text=Continuous%20feedback%20helps%20you%20evolve,%2C%20product%2C%20and%20website%20experience.&text=User%20feedback%20provides%20actionable%20insights,significantly%20impact%20the%20user%20experience
McKinnon, T. (2023, May 10). How Starbucks is using mobile apps to significantly increase sales. Indigo9 Digital Inc. https://www.indigo9digital.com/blog/starbucksmobileapps
Neusesser, T. (2024, September 6). A/B testing 101. Nielsen Norman Group. https://www.nngroup.com/articles/ab-testing/
Simkus, J. (2024, June 10). Between-subjects design: Overview & examples. Simply Psychology. https://www.simplypsychology.org/between-subjects-design.html
Starbucks. (n.d.). Nov. 14 is Starbucks Reusable Red Cup Giveaway Day. About - Starbucks. https://about.starbucks.com/red-cup-giveaway-day
Trappe, C. (2024, February 16). Research incentives 101: Setting appropriate compensation. Virtual Incentives. https://www.virtualincentives.com/research-incentives-101-setting-appropriate-compensation-2/

© 2024 Copyright by Gabriela Gonzales. All rights reserved.