Foursquare’s Taste feature is a great example of solving a non-obvious product problem — it challenges you to think beyond familiar domains like e-commerce or banking.
Foursquare was a location technology platform that helped users find the best options nearby — restaurants, nightlife, shops, and experiences — based on their current location and personal preferences. It is an example of a product that was ahead of its time, combining mobile GPS capabilities with social networking to provide personalized recommendations.
The actual job Foursquare set out to do was not just to show places on a map. It was to make each customer's experience unique and personal, building a strong bond through relevant, timely, and context-aware suggestions.
This lesson uses Foursquare’s “Taste” feature as a case study to teach you how to think about product epics, user stories, and tasks when the product domain is unfamiliar. If you can solve for this, you will be ready to tackle any unknown product scenario.
Foursquare’s journey: from check-ins to personalized discovery
Foursquare began in 2009 as a joint venture between Dennis Crowley and Naveen Selvadurai. It was the successor to Dodgeball, a location-based social network acquired by Google and replaced with Google Latitude.
The key insight behind Foursquare was that mobile devices could be used not just to connect people but to provide lifestyle recommendations based on location. The rise of smartphones with built-in GPS enabled accurate location detection — a foundation for personalized local search.
Initially, Foursquare focused on social check-ins, awarding badges to users for visiting locations. Over time, it evolved into a local search-and-discovery app with personalized recommendations driven by a user’s check-in history, browsing behavior, tastes, and social network ratings.
By 2017, Foursquare had converted its vast user-generated location data into a sustainable revenue stream, serving 3 billion visits per month across 105 million venues worldwide, with 25 million active users.
The core challenge was keeping users engaged and continuously feeding data, which was critical to maintaining the value of their location intelligence platform for advertisers and businesses.
The challenge of personalization: the “Taste” feature
The heart of Foursquare’s differentiation was its “Taste” feature — a mechanism for users to define their preferences across food, environmental factors, travel styles, and more. This feature aimed to make recommendations deeply personal and relevant.
A new user could select from a predefined list of tastes — for example, “live music,” “vegan,” or “outdoor seating.” These tastes informed the recommendations they received, so if “live music” was selected, the user would see venues known for live performances.
This personalization went beyond simple location proximity or popularity. It accounted for individual preferences, time of day, and social signals to tailor suggestions that felt uniquely relevant.
Defining the theme: personalization as the core experience
The theme for Foursquare’s Taste feature is clear:
Develop features that provide a unique and personal customer experience by delivering recommendations based on individual tastes and preferences.
This theme captures the product’s intent to move beyond generic location search to truly personalized discovery, increasing user engagement and loyalty.
Formulating the epic hypothesis statement
The epic hypothesis translates the theme into a testable product vision:
“The Taste feature helps users receive personalized recommendations of places based on their stated preferences and past behavior, increasing engagement and satisfaction.”
This hypothesis guides the team to focus on capturing user preferences and using that data to improve recommendation relevance.
Writing user stories from the user perspective
User stories break down the epic into actionable chunks, always phrased from the user’s point of view:
-
As a user who has added tastes, I want to receive recommendations tailored to my preferences so that I discover places I like.
-
As a new user, I want to select my tastes from a list so that my profile reflects my interests.
-
As a user, I want to update or remove tastes at any time so that my recommendations stay relevant.
-
As a user, I want to search and filter recommended places based on my tastes and other criteria (e.g., time of day, location) so that I can find the right place quickly.
-
As a user, I want to provide feedback on recommendations so that the system learns and improves over time.
Breaking down tasks to achieve the epic
Each user story translates into tasks for design, engineering, and data teams:
-
Design and implement the “Taste” selection interface: create UI components for users to browse and select tastes.
-
Build backend APIs to store and retrieve user tastes: ensure preferences are saved securely and efficiently.
-
Integrate taste data into the recommendation engine: modify algorithms to weigh user tastes alongside location and check-in history.
-
Develop search, filter, and sorting options: enable users to refine recommendations interactively.
-
Implement feedback mechanisms: capture user ratings and comments on recommendations to improve personalization.
-
Test the system end-to-end: validate that taste data impacts recommendations as expected and that users can update tastes smoothly.
Visualizing the relationship: theme, epic, user stories, and tasks
A concise diagram helps clarify the hierarchy:
-
Theme: Personalization through tastes
-
Epic: Taste feature delivers personalized recommendations
-
User Stories:
-
Select tastes
-
Receive tailored recommendations
-
Update/remove tastes
-
Search and filter recommendations
-
Provide feedback
-
-
Tasks:
-
UI design for taste selection
-
Backend APIs for taste data
-
Recommendation algorithm integration
-
Search and filter implementation
-
Feedback capture system
-
Testing and validation
-
-
-
What this exercise teaches you about product management
Most PMs find familiar domains like e-commerce or banking easier to write user stories for because they understand the user context intuitively. Foursquare’s Taste feature is deliberately non-obvious to push you beyond that comfort zone.
The actual job is to extract the underlying customer problem and break it down into manageable, testable components — no matter how unfamiliar the domain.
If you can do this for a location-based personalized recommendation system, you can do it for any product.
Reflecting on Foursquare’s product experience today
Looking back, Foursquare faced challenges in engagement and interactivity. Unlike platforms such as Instagram, which combined content and social interaction deeply, Foursquare’s experience was more transactional — a platform to find places rather than a community to explore and share.
This limited sustained user engagement, especially as other mapping and social platforms integrated location features.
But the core idea — leveraging location data and user preferences for personalized discovery — remains relevant and continues to inspire products today.
Test yourself: Defining epics and stories for a new feature
You are the PM at a local discovery startup in Bangalore. You want to build a “Mood” feature that recommends places based on the user’s current mood (e.g., relaxed, adventurous, social).
-
Define the theme for the Mood feature.
-
Write an epic hypothesis statement.
-
List 3-5 user stories to deliver the feature.
-
Break down the tasks needed to build the MVP.
You are the PM at a Bangalore-based local discovery startup. Your CEO wants to build a Mood-based recommendation feature. You have two weeks to prepare a product definition.
The call: How do you define the theme, epic, user stories, and tasks to ensure clarity and focus?
Your reasoning:
Where to go next
-
Learn how to write effective user stories: User Stories and Acceptance Criteria
-
Explore product discovery techniques: Product Discovery Frameworks
-
Understand prioritization frameworks: Prioritization Techniques
-
Practice breaking down complex features: Epic and Feature Decomposition
PL alumni now work at Flipkart, Google, Razorpay, PhonePe, Swiggy, Amazon, Microsoft, and 30+ other companies.