Clarifying questions and assumptions:

1. Which geo are we solving for? I would assume its in the context of Instacart in USA/Canada
2. Are the problems new or have always existed? Assume that the problems have existed and not new

I will first start with mapping the customer journey. Then make assumption about what might be biggest pain points.

A user journey post placing an order looks like and what the pain point associated with each might be

1. Shopping for order begins and incases where order replacement is needed or cannot be found one needs to chat. Some pain points associated with it would be (a) There are many items that need a replacement (b) Takes investment of time. One has to continuously chat with the shopper to decide which items would need replacement
2. Order receival by person: In this case, the pain point might be that the order is not dropped at the right location or it takes too long to get the order
3. Quality of items/check against order that was placed – In this case, the pain points might be (a) that items received are abt to expire (b) items are rotten like fruits (c) it states online that item was received but in actual the item was not received

1. Instructing employees on ensuring items being delivered are fresh vs. not fresh. This can be done by building in-product notifications or training to shoppers that they should be selecting fresh items for employees. Instacart can also take a look at data if this problem is specific to a certain set of instacart shoppers and cater these reminders/notifications to them
2. In certain cases, like milk/fruits/non-frozen foods, letting ppl enter by when they expect to consume so that the freshness can be gauged
3. Also, Instacart should look at the data if this problem is specific to a certain set of grocery shops where there might not be as many fresh items. In that case, one should let those stores know and work on building partnerships with other stores

1. In measuring success, if the measures put in place are working or not one should look at the consumer ratings and if there is improvement in that. Today, Instacart does provide an option to rate what they liked abt the experience and one such thing to grade on are Quality of items.
2. We should also look at how the Instacart shoppers who are getting notificaition engaging with it. Are they engaging more with these notifications?
3. Similarly from consumer end we should look at how many times ppl are entering notes on when they expect to consume the item etc.

Improve the worst customer experience post a successful order placement on Instacart

• What is our goal?
• Why are we improving the post-order experience?
• did we hear any complaints from customers that we want to fix?

1. End users
2. Supplier or grocery retailers
3. Shoppers – pick up and delivery person

Related orders

1. Wrong order delivered
2. Deliver to a wrong address
3. Quantity is incorrect
4. Damaged items
5. Bad packaging
6. Quality is bad
7. The order did not arrive on time
8. Unable to know the exact order status
9. Items are not in stock after the order placed
10. Order not at all delivered

Related to Shopper or delivery person

1. Different person deliver the order
2. I am not able to trust the shopper
3. A person is not ready to deliver to the address at the mentioned location
4. A person is arguing and asking for the extra tip
5. A person is not ready to deliver an order, or he is a fraud – For example, he is taking an order and not responding

1. Do not deliver the out of stock items and do not do any updates and refund the remaining money  [Efforts: L and Impact: L]
2. Proactively send a notification and inform them we will not deliver OOS items [Efforts: L and Impact: L]
3. Proactively send a notification and ask them if they would like to update the order with anything else instead of OOS items. [Efforts: L and Impact: M]
4. Take input from users proactively while taking an order [Effort: M and Impact: H]
5. Send a question with in-app suggestions of another store as an alternative place to pick up same item or another item from the same store [Efforts: M and Impact: M]
   1. Use ML and based on the data come up with the solution.
6. Start a conversational channel between the end user and the bot to pick if we could get the end user’s thoughts if they need alternative items, the same items from another store, or any additional items.  [Efforts: H and Impact: H]

Before I get into the problem approach, I would like to seek a couple of clarifications on the context and the situation

Instacart helps deliver and pick up products from your local retail stores. Although Instacart is famous for grocery delivery, it’s venturing into electronics, pharmaceuticals and other types of retail store deliveries.

Clarifying scope

1. Is my role to design and work on fending off a worst customer experience as an insta cart product manager –Yes
2. Are there any resource / market constraints that I should consider– No?
3. Can I work on the problem scope in US market and expand later based on the culture of other regions– Yes
4. Instacart works in collaboration and has an experience built into Individual merchant apps, do we want to build this experience into Instacart alone and not merchant apps – >Yes

Business Goal – Our business goal is to improve the customer experience by mitigating some of the problems after a customer place an order in Instacart which will drive user engagement. A happy customer is always expected to come back and our goal is to drive engagement. Please let me know if I can proceed further.

Users:

1. Customers
2. Instacart deliveries
3. Merchants

I would like to focus on the customers as we would like to solve customers’ delivery experience in this scenario

I would like to approach the above problem by first, identifying customer needs, I would like to then look through the most prioirtized needs, secondly, think of solutions that could be implemented and recommend a solution with a metric to see help us meet our goal.

Ideally, It would be great to look into the data and see which are is being impacted the most and since we do not have the data here, I would like to go through each of these needs and prioirtized based on my customer instinct, I would like myself to stand in the shoes of a customer placing an order

 Customer Needs

While all the above is a priority to the customer, I would like to focus on the damaged items delivered as generally that leads to a bad customer experience and refrains them from leveraging the service again, customers expect quality in the product being delivered.

Here are some of the solutions that come to my mind for avoiding damaged items being delivered

Of all the solutions above, I would like to recommend the 3rd option where we are being proactive and addressing one of the issues with deliveries.  We can do a A/B testing on the special instructions surfaced vs not.

North Star metric

1. # Of issues reported on damaged Items – Take a baseline metric and observe improvement with the ML model in place.
2. DAU – Secondary metric to track is the increase in the daily active usage