Increasing In-App Orders On The Starbucks App

Introducing a new feature to increase orders via the Starbucks® mobile app

View Project File

Starbucks is an international chain of coffee shops with a popular mobile app – the Starbucks app. App users can customize, place orders, and pick up from stores without waiting in line.

“The Starbucks® app is a convenient way to order ahead for pickup, scan and pay in-store, and customize your favorites. Rewards are built right in, so you'll earn Stars towards free drinks and food on your purchases.”– Google Play Store.

The coffee shop space is immensely competitive and the company faces competition from other large international chains (McDonalds, Costa Coffee, etc.), regional coffee chains, and well as independent coffee shops. Many competitors have apps of their own with their own sets of unique features.

Let’s assume that Starbucks is interested in increasing orders via the Starbucks® mobile app. As a Product Manager, I will define a new feature that will help Starbucks accomplish this goal.

Assumptions

• We validated the problem and determined that building new features is a business priority

• Geography - US, major metropolitan cities (>100k population)

• Business Goals - Features selected for development will prioritize increasing order frequency and order value for purchases via the Starbucks® app.

• Out of Scope – Payments, Starbucks Rewards, and general app operations (login, settings, etc.)

User Flow – As-Is

Before we begin to introduce new features, we need to examine the current user flow – How are customers currently making purchases on the app?

To identify potential new features we can introduce, we need to conduct user research to understand the challenges and potential improvements to the app.

User Research

To understand the current user experience, we can engage in a discovery process to examine potential user needs and pain points.

User Interviews

Interviews were conducted with 20 different people and observations were made during order pickups and interactions with staff, to gain insight into the user experience from start to finish (App Open to Order Pickup).

Questions

We asked a series of questions, focused on the app experience, specifically around placing an order.

Current App Usage

• How have users been placing orders using the Starbucks app?

• How often do users place orders on the app?

• What situations prompt users to place an order on the app?

• Are there any features or capabilities that users find particularly useful?

• On a scale from 1–5, how satisfied are users with the ordering process?

Improvements

• What is the measure of success when it comes to placing an order?

• What challenges/inconveniences do users experience while ordering?

• How are users currently addressing these challenges/inconveniences?

• What could we do to make the ordering process better?

• If users could change/add one feature on the app, what would that be?

Note: This is not an exhaustive list of the questions asked but rather the key questions of interest.

Feedback

The feedback gathered from the user research was categorized and organized into groups. Furthermore, we validated the feedback by reviewing 100 recent Google Play Store reviews to identify the frequency of associated issues in reviews related to “orders.” We can use the reviews to add context to the feedback based on the frequency of these types of issues – Are other users also experiencing this issue? Feedback could be further validated using customer support data but we do not have access to that at this time.

Note: We are considering issues related to in-store purchases to be out of scope since we are focused on enhancing the mobile order process.

User Personas

Target User  - Barry, 28, Professional

Out of all the personas, this user is likely to have more predictable habits, consistent spending, and moderate order values. While the other user personas may have higher order values their spending is less predictable and subject to change based on their schedules. Additionally, the “scheduled pickup” featured does address the pickup concerns of the other personas.

Potential Solutions

Feature	Description	Impact	Effort
Scheduled Pickup	An order scheduling order option to select a pickup time when placing an order. Pickups can be scheduled within a certain period (Ex. order for pickup 30 minutes in advance).	4	3
Accurate Pickup Estimate	An order time tracking system that would monitor the average order preparation time at each store, the volume of orders coming, and the average pickup time, to calculate accurate estimates for each store.	4	5
Favorites Edit	An option to quickly edit items in “Favorites” so they can be added to orders. The core order process remains the same.	2	2

Note: Grading on a 1-5 scale with 5 being the highest impact/effort and 1 being the lowest.

Prioritization

Scheduled Pickup - Prioritizing this feature because it simplifies the uncertainty around the pickup process for the target user, allowing them to plan and manage their time more effectively.

• The expectation is that this feature will increase order frequency and satisfaction for users who would otherwise be turned off due to pickup delays and inaccurate estimates.

• Users can now schedule orders for pickup throughout the day instead of placing orders while en route to the store. Giving them the peace of mind that their orders will be ready in case they are busy commuting, driving, or engaged in other activities.  

• To prevent overwhelming, stores with excessive orders, this feature would be limited to same-day orders made 30 min before the store closes.

• Scheduled pickup also increases the “Lead Time” for the store that must prepare an order.

  • For example, at a busy store, preparing an order with 30 min of advance notice is easier than preparing an order with 5 min of advance notice.

  • Store employees can manage their time better when they have a clear knowledge of when a customer might arrive for pickup. It also allows the employees to prioritize tasks more effectively.

Accurate Pickup Estimate - While it would be great for users to know the wait time for their order, it would be challenging to implement.

• The time it takes to prepare a certain order depends on the volume of mobile and in-store orders coming in, the number of employees on staff, and the complexity of a specific order.

• The current system is not accurately estimating the pickup time, which means a detailed investigation may be required to properly diagnose, identify, and address the problem.

• It may require a redesign of the current system used to track order processing times, which could be complex and time-consuming.

Favorites Edit - The ability to edit favorites simplifies the process of placing an order but does not resolve the user’s challenges with speedy and timely pickup.

Measurement

Penetration/Usage – “Scheduled Pickup” Feature Users

• % of target users that initiate at least one instance of “Scheduled Pickup”

• In comparison to the general user base (pre/post feature)

  • Avg order value of feature users

  • Avg order frequency of feature users

  • % of users that cancel scheduled orders

• Mentions of this feature on social media

Supply/Side – Starbucks Stores

• Number of scheduled orders vs. normal mobile orders

• Compare tipping behavior (amount & frequency) to on-demand orders

• Interview store managers to determine if these scheduled orders produce better consistency in order preparation and order management

Course Correction

The major assumption is that this feature solves a significant problem related to order pickups

Expected feature usage - >10% of the target user base

If usage is <10%, analyze positioning, conversion, and drop-offs, throughout the funnel.

Expected increase in order frequency per week - +1 additional order per week per user:

If frequency doesn't increase, reconsider the hypothesis, and gather more feedback.

Risk Factors

• Stores may not be able to guarantee orders will be ready at a scheduled pickup time in case of unexpectedly high order volumes.

• The feature is not utilized as much as predicted by the target users

• Increase in cancellations due to this feature.

• Decrease in customer satisfaction resulting from unwanted charges due to this feature.

• Increase in support requests regarding refund requests due to this feature.

• Decrease in repeat usage if users learn to plan more in advance.

• No reduction on order pickup times – customers still waiting for pickup.

Risk Mitigation

• Allow stores to limit scheduled orders if total orders exceed a preset limit.

• Select the test group most likely to adopt (via surveys, interviews, etc.) and run a pilot with MVP to confirm assumptions regarding these recurring orders.

• Monitor the volume of cancellation requests related to this feature.

• Survey users to monitor satisfaction levels prior to and after using this feature.

• Monitor the volume of support requests for refund requests related to this feature.

• Provide additional incentives for users using this feature (discounts, promotions, etc.)

• Prompt participating stores to prioritize recurring orders over regular mobile orders.

Summary

We will be introducing a feature to implementing same-day “Scheduled Pickup” to allow users to order pickups, making pickup times more predictable and convenient for consistent usage

• Validate the solution with the target audience, confirm assumptions and hypotheses with primary and secondary data

• Validate solution with internal stakeholders (Store Ops, Sales, Marketing, etc.)

• Monitor usage patterns from existing users and compare to other cohorts, pre/post feature launch to understand changes in behavior

• Create incentives to reinforce the desired behavior (discounts on scheduled orders, stores prioritizing scheduled orders, etc.).