Amazon Fresh - Build a Product Flashcards
What is the framework for defining a product for Amazon fresh?
PAUSE - take 1-2 minutes to write down your thoughts
1. Amazon Vision
2. Brainstorm 3 user segments we can pick from
3. What needs and pain points might these customer have?
4. Feature ideation 2 - 3 features and prioritize
5. Prioritization Framework
* Feature Vision
6. How does the feature work
* Frontend
* Backend
* Business teams involved
7. Tradeoffs
8. Metrics
What is the Vision of Amazon?
“serve consumers through online and physical stores and focus on selection, price, and convenience.”
Customer Segmentation: For defining an Amazon Fresh product, what 3 customer groups should we use to define a customer segment?
List out 3 user segments:
-
Stage of life -
- Kid
- Young Adult
- Mature Adult
- Elderly
-
Profession -
- Student
- College Student
- Young Professional
- Mature Professional
- Retired
-
Cost Sensitivity
- Highly cost sensitive
- Moderately Cost Sensitive
- Low Cost Sensitivity - high budget
Needs/Paint Points: What are the needs & pain points for an Amazon Fresh customer segment?
- Convenience
- Broad Selection
- Price
- Time to receiving goods
- Ease of purchase
Ideation: What are 3 solution ideas?
-
Whats my community loving?
- Featured recommended product that people in their community is loving.
- “What’s everyone else eating”
-
What’s for Dinner - by Amazon Fresh
- Recommended Recipe based on the items in your cart.
- “Have Amazon help you with what to cook”
- “Create this meal again”
-
Auto Essential by Amazon Fresh
- Perishable food reminder
- “You might want to replace your essentials”
What is a good prioritization framework for choosing the best feature to work on?
Prioritization:
- Impact
- Confidence
- Effort
Whats the framework for describing the solution - “Whats everyone in your community eat”?
* Focused recommendation with a deal”
-
Front end
- How will the user interact with it?
- What will it look like?
-
Back End
- What will we need backend to support it?
-
Business Teams Involved
- Vendor Management
- Deals Team
- Buying Team
- Tech, Product Management
- Data Science
What kind of vision can we develop for Amazon Fresh - focused recommendation?
“Whats my community loving” by Amazon Fresh
Taking the Recommendation section to the next level
List some tradeoffs that we might face with an Amazon Fresh focused Recommendation.
- Very data intensive, especially if done by region
- May take valuable real estate on the Amazon Fresh homepage
- Evaluate results - don’t want to get bread constantly
How would you design the Front End for “Focused Recommendation” ?
Draw it out!
- Focused Recommendation - simulate the experience of a friend recommending a new product to you.
- Showing you one - three recommendations, and visible reviews.
- Ideally we have a deal attached to it - low the barriers to trying something new.
- Double pane that shows one of two options for “Why” its a recommendation..
- “Because you bought this” or “Because this is your repeat item” .. why don’t you try this?
- This is “Trending in your community”, you should try this
- Recommendation algorithm that takes the “best fit” recommendation based on “Trending” or
“Because you repeatedly buy this”
How will the Back End work for Amazon Fresh “Focused Recommendation”?
- Back-end ML Algorithm
- Takes into account:
- Number of reviews on a product
- Number of reviews on a product in your general community
- Geography - 1-6 zip codes
- Number of purchases in your general community
- Product type
- Product Price
- Promotional?
- Prioritize newness
- Takes into account:
What business teams are involved in bring in the Amazon Fresh “Focused Recommendation” to life?
- Data science team
- Amazon Deals team - responsible for negotiating deals for certain products with big brands
- Product Management
- Buying Team
- Vendor Managers - sales, promotions, customer experience
- Do we have the inventory to support this product being “recommended”?
- Do we like this recommendation? Content Management
How might we test this feature?
- Prioritize a major urban area for an A/B/C test
A of the original homepage (our control)
B of the “Trending in your community recommendation”
C of the “Because you repeatedly buy this?
What metrics should we focus on?
- Click-through rate
- How many customers are we seeing click the recommendation
- Add to Cart conversion
- How many customers end up adding the item to the cart
- Purchase Conversion
- How many customers end up purchase the item recommended
- Cart Abandonment Rate
- What percentage of customer abandon the recommended time in the cart