Our first meetup in San Francisco is coming on Jan 23, 2020. Learn more.

If you were changing the AmazonFresh experience what are some are your ideas and how would you measure them?
+3 votes
in Metrics by (717 points) | 680 views

2 Answers

+12 votes

I have to cheat a little, I have actually never used AmazonFresh before. From my understanding of AmazonFresh, I believe it’s a part of the website where you can order fresh groceries delivered to your door. I believe there’s some sort of brown bag (?) that you have to fill up for free delivery. I also think that you are able to set schedules for delivery of groceries.

If I were to change the AmazonFresh experience, I would have to first define why I am changing it? is it to get higher DAU? Higher revenue? Higher returning users? I’ll choose one at random, I want to change the experience to get higher returning users. I want to go through some user stories and identify potential issues of why users may not come back to amazon fresh.

A user goes to amazon fresh because they want to buy groceries. Chooses from a list of items that they potentially want. Aim to hit the minimum for free delivery, checkout, and reserve a window for delivery. The user waits for the delivery after work hours and obtains the groceries.

There are some potential issues and again please let me know if these issues may already be solved for.

– Users may want a simpler process to identify key items they want to purchase
– Users may get items that are not as fresh as their standards (ie milk with a date that is expiring soon)
– Users may not always get upt o $50 and have to pay for delivery cost.
– Users may miss the normal grocery shopping experience of choosing ingredients themselves, and have ingredients inspire them
– Users may have have purchased the wrong amount of items, or have limited understanding and have to go out to the grocery store again regardless

I want to identify some potential solutions, but keeping in mind this is to improve customer experience so they will come back and shop on amazonfresh long term.

1. Amazon can create a shopping list widget for people’s desktop, or an app on their phones and integrate with all major personal assistants (google/siri/alexa). The purpose of this is to allow people to build a grocery list that amazon will keep track off on their website.

Additionally, Amazon may want to present options when user logs in, so if someone is like I want green Onions, well here’s a list of potential options. Users will click on a “saved list option” and see their list presented with options that they can click on.

2. For users that have a negative experience on freshness, we should immediately follow up with plans on redelivering their items. Amazon is historically really good at customer service, but we may want to reserve some delivery space everyday for these special complaints in order to service customers who had a complaint.

3. I think the best way to solve for this is 1) continuously explore ways to lower the minimum cost and 2) help users bunch together items to hit their target minimum. To focus on number 2, once amazon has built a history (and they can with their own purchase history outside of amazon fresh), we can create a widget that reminds users to purchase certain items. For example, if hte last time you pruchase shampoo was on amazon 2 months ago, we can present a “did you want to buy more shampoo as well?) This will all be presented on the shopping cart page to help users hit their minimum.

We can also leverage amazon shopping list widge and app to present a easy way for users to understand how far/how close they are to the free shipping.

4. I would build a recipe system. Build as part of the app, and web surfing experience a calendar plan of things you want to cook and want to eat. This way, amazon can suggest ingredients to you that are both seasonal, and delicious. Partner with some major chefs to create great recipes that allow users to both search for suggestions, but also enter their own. From there, amazon can create a shopping list for them and decide which day makes the most sense for delivery.

5. I think this issue can best be solved operationally by letting users know where the nearest whole foods is and allow users to easily either pick up missing ingredients or return items that are inappropriately purchased. It’s sometimes hard to avoid last minute situations and I believe having the store front will really help with the experience.

I’ve identified several ways to make amazon fresh a much more powerful experience that both relies on user engagement and user purchase history to make more powerful. I believe this will help customers use more of amazonfresh in the future.

I would prioritize releases based off of impact, goal, and cost

For 1, shopping list will have a high impact and might actually help organize the users life easier. It hits our goal because it builds the list over time and allows a user to leverage amazon fresh by planning ahead. The cost should be medium to high depending on how many apps and integration to come out with. I would first integrate with amazon echo devices and go through the mobile apps.

2, I believe this has high impact, reaches the goal, but may have high operational costs. To be honest, I think Amazon probably already does this.

3. High impact because free delivery is probably a primary concern for most users. High goal, and medium cost as a recommendation system is probably built in and we just need to surface the right information to the user to help build out their shopping cart.

4. Recipe system can be complex or simple in nature. I think it will hav ea high impact, high goal, and also potentially a high cost.

5. This one is harder to estimate. I think it will have high impact, high goal, but operationally may cost a lot as well. It’s really hard to manage consumer behavior during last minute times but I don’t necessarily think that this issue is keeping customers from using amazon fresh in the first place. I would run more experiments and understanding more about customer needs.

I would build 1,3,4 as first release. KPIs I’d be most interested in is:

1. how often users engage with these new systems
2. How often are customers shopping on amazon fresh, before and after
3. Segment by each feature, were customers before the change shopping more after the change per key feature and also combined?
4. Do customers who hit the free delivery churn less than customers who don’t?
5. How often are customers shopping on amazon fresh per week?

With the key KPI being customer shopping growth on amazon fresh because we are optimizing for returning customers being happy with the system and having amazonfresh augment their normal grocery experience.

See less
Great answer!
0 votes

Great answer.One thing, though:3. Segment by each feature, were customers before the change shopping more after the change per key feature and also combined?I would do an A/B test instead of looking …See more

by (15 points)
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
To avoid this verification in future, please log in or register.
Your name to display (optional):
Privacy: Your email address will only be used for sending these notifications.
To avoid this verification in future, please log in or register.