Interesting one. So I think any problem is easier to solve by understanding the problem and then thinking about why we may want to do this and for whom. So let me first understand this a bit.

1) We want to design what type of product? Hardware / software / service? (don’t matter)

2)ok I think I will lean to software product. But are we designing this just for a particular store or any store? (any store)

3)can we do this just for US stores since I am more familiar with US market? (sure)

4)Alright any do we have any guidance on why we are doing this? I mean, has leadership expressed any particular reasoning? Do we have any specific user data based on research, etc.? (yes the data we have points to making a product but the information doesn’t help us determine what we need to build so as a PM we want to see if you can help solve this).

5)So is it fair to say that customers find a lack of product that helps them in shopping for a wine? Yes that’s what the queston states.

6)ok….so given that I want to think of who my users might be a)Wine affeniados (which I don’t think I expect in a grocery store unless this is a grocey store in a town that’s a wine country but that’s an edge case at least in US) b)budget shoppers c)shoppers who like to casually drink wine but have some choice and knowledge. Now I realize that I could also have shoppers who are every weekend drinker or weekly drinkers and those would be more frequent if you will but let’s go with 3 types for now. I just assume that affiandos would be more so going to wine depots since they need ample selection and since grocery stores don’t carry many wines it prob attracts more casual drinkers who are seeking for reasonable prices. I realize this is sort of a combo of b and c but I think they are likely the same.

Interviewer: interesting. Are you sure you want to think of it this way?

Me: Yes for our purposes it makes sense. In real life, however, I would wnat to coduct additional research to prove my hypothesis.

Interviwer; ok

Me: So since my user is interested in wines and seems to have some choices, I may wnat to design something that helps the user. I also think that my product would be much more adopted if I throw in a bone for the grocery stores. Here is what I am thinking ….

1) Grocery stores always love more shoppers

2) Wine shoppers who have some knowledge are seeking to grow their knowledge and prob looking for the right wine for themselves

3) Wine shoppers may get tired of wines that are carried in their nearby grocery store.

I think #3 and #2 are more important and I think #1 can be solved potentialy if thats what we focus on. + I don’t think #1 is well algined to the #2 and #3. So let’s focus on #3 & #2 for a moment.

Give me a min please…

Solutions:

2a) Maybe my app allows me to scan a wine and it tells me information about the wine. Not just the info I find at the back of the label but also other important things such as sugar content, months fermented, something about the region of that particular wine, notes (on the note and pallet), how spicy it is if it is, dry vs sweet, acidity level, tannins, etc.

2b) The app should start recognizing my taste and have a section where it tells me waht my taste is based on data. When my taste changes slightly based on my rating of a wine, it will add to the timeline and note what taste has changed and what has remained the same. I think the app should also consider seasonality that might be affecting my tastes. So maybe there are two sections for timeline (1 for summer time and 1 for wwinter time or 1 for white and 1 for red).

2c) App will suggest me what type of wine to get based on my taste. Some people drink red during winter and white during summer so identify those things and prompt me “this is about the time you do white wines. You last had a taste of this for white wines and this store has this one. Wuld you like to try one of these? On a scale of 1-10, we would suggest you try this 1st, this 2nd, this 3rd. Let’s try those and see.”

3a) Wine can suggest me stores that have wine that I may like more. This means, we are going to have to make this a 2 sided marketplace app. And grocery stores would have to subscribe to this app. There is interest in subscriving for grocery stores since this is a marketign channel for store to get more customers in their shop.

3b) Since we are now making this a 2 sided marketplace, I think customers should be able to suggest what wine they would like for stores to carry. Maybe there is a db through API that we can validate hte wine exists out there. This way stores can have demand data before they choose to carry such wines. However, there might be legal and regulatory challenges involved with this one. I have haerd from a wine store in Arakansas, that laws govern what specific wines a store is allowed to carry. It may not be true but if true we would have to incorporate some API to check if that wine can be carried for that area and inform the customer that the wine can’t be added as an interest in it.

2d) I am going back to a solution for #2 pain point since I just thought of it. I think customers should be able to suggest a wine to customers of similar taste with just a simple click and a broadcast message. Also, I as a customer should be able to add to my friends list so then I can specifically suggest a wine using the app to a friend of mine.

I think these are great amount of solutions but to build something I am going to have to consider any constraints on resources, time, money, etc. Do we have any?

Interviewer: I am glad you asked. We want to have it ready for Thanksgiving so you have 3 months fron concept to launch.

Me: So we are going to have to choose something small and low hanging fruit if you will. I will want to build 2 or 3 features or else the app won’t be too valuable. I think we should go with 2a, 3a, 2d. I think these directly address immediate needs my customer of choice may have such as “tell me what store has the wine I am going to like, I don’t lknow what I want since I think I am running out of choices, and I have osme knowledge so I do watn to know more about specifics of teh wine I drink”. The challenge with some of tehse solutions is being able to recognize a specific wine based on picture + some wines have the same name and label but depending on year of production taste is slightly different so app is going to have to somehow solve this problem.  + what if that wine is not found?

I think there potential frustration points here so immediately after thankgiving I am going to want to see if the  # of users who started using it before thanksgving or during thanksgiving  is stayed flat or gone up after thanksgiving. But it may go down just because after holidays people are now trying to cut down on calaroies before x-mas heavy dinner and parties. I am going to want to check that # again right before and after x-mas. I would wnt to also see if the same users are logging in again and again or not so maybe # of sessions of these users.

OVerall, I think it’s imp to make an app that’s a 2-sided marketplace and an app focused on providing information about wines to my customer so that my customer is given guidance and help as to what wine to buy from the store.

Clarifying questions

Is the app going to be used for wine shopping only or for grocery shopping in addition to wine? – Wine only

Is this any grocery shop or a particular grocery shop? – Any

User groups – I would say there are 4 distinct groups here :-

1. Connoisseurs
2. Frequent low-end wine drinkers
3. In-frequent occasional/event-only wine shoppers
4. Gift shoppers

I do not think users in group #1 will be shopping for thier wines in grocery stores. They will go to speciality wine dealers. #3 & #4 do not promise a large market, so I will focus on #2 for the app.

Pain points/User persona of wine shopping for user group #2 :-

1. Value shoppers
2. Weekly or bi-weekly purchases of wine
3. Not brand loyal
4. Willing to try new wines but do not have enough information on how to choose

Success with group # 2 will depend on increasing their frequency & quantity of wine purchase by making them more engaged and helping them make informed choices.

Features – here are the features I would consider for the app

1. Guided shopping experience – Ask them a series of questions (similar to a store rep trying to choose a wine for them) and suggest the best value wines within their budget based on their answers.
2. Account creation –
3. Ability to remember their previous purchases which they liked to suggest similar wines in the future
4. Ability to browse & notify the best ‘new’ wines for the week
5. Ability to browse & notify deals of the week
6. Scheduling wine tasting events at the store

Based on this, there are some low hanging fruits which can be implemented first like ‘deal of the week’, ‘browsing new additions’.

These should be followed by features which are of High importance to users or to business success with medium to high effort like “Guided shopping’ & ‘Remember previous wines to give new recommendations’

Clarifying questions:

What kind of grocery store- chain or standalone? -> let’s say -> Facebook is designing for users to help browsing and purchasing wines at any grocery store

is it targtted to grocery stores be online and offline – both

Wine shopping new feature of the app? Yes

Buyer and seller – any one to target in particular – no

Constraints – time, money, resources – no

Goals:

Mission – keep people socially connected

Vision – make facebook a one-stop shop for people to use the platform to discover their connections and whatever that interests them

Why is Facebook interested – just like facebook, wine brings people together – whether it is going for wine tasting, or bringing them at home for a nice dinner party, it keeps the social engagement going. So, they follow the same theme of building social connections – whether it is while drinking or on the internet. Facebook wants to help its users drive more experience when it comes to buying wnes – to identify the right wine for onself by building on its core strength – social connections and drive towards the company’s vision

Possible canniblization – there are wine groups on Facebook where people do talk about wine tasting events, different kinds of wines, online wine shops. Having a sepcific feature gives users a unique, centralized experience where they can share different facets of wine buying/drinking, thus creating more engagement

Business goal of this feature – more engagement (short term), retention (med), and monetization (long term). For the excerise , I’ll focus on the engagement

User segments:

Buyers:

1. Adults having limited to no-knowledge about what wine to buy/drink
2. Adults who with average knowledge/medium level of drinking wine
3. A new restraunt in town planning their wine collection
4. Adults who are pros in wine tasting and buy wine at the store

Sellers:

1. New wine sellers promoting their wine in store
2. Wine experts wanting to help people with wine related questions
3. A new wine shop promoting its wine collection

Prioritized – adults having average knowledge/medium level of drinking wine

User persona

John Doe, works in tech in SF. Lives with roommates and likes having a glass of wine at the end of a busy day, budget shopper

Wine knowledge level – knows wine catgeries – red, white sparkling. Likes red wine. Has some basic idea (but is not a pro) of how taste differs. In other words, develops an understanding of wines by buying different wines at a grocery store (neednt be the same store everytime)

Purchase frequency – a bottle in two weeks

Journey

1. Researches for wines online – different websites, talks to friends, browses different stores
2. Looks at the area from where the wine is from, read reviews, ratings, price, store ratings, picks one that sounds like a good choice
3. Goes to the grocery store, may browse some other wines but and buys the one he wants to drink
4. Tastes it and makes an likeness/dislikeness towards it
5. Based on like or dislike, reviews the wine in his mind or online and makes a re-purchase, shares it with friends

Pain points

1. Since everyone has a specific taste, reviews and ratings dont help much in judging the taste of the wine
2. Budget
3. Since John is buying new bottle in two weeks and likes experimenting, he needs more variety to experiment on

Prioritized – # 1 and #2 – reason – tastes and reviews can be very subjective (since everyone has a different taste and opinion), so , it becomes a critical pain point in addition to the budget. Addressing these pain points will drive more engagement, which is our business goal

Solutions

1. Wine Shop directory – list of online and offline wine/grocery stores based on availability of wines based on the history of wines tasted, bought, location, review scales, budget
2. Wine match predictor – predict the wine based on the mood, food accompanied, personality traits, budget
3. Wine Expert Platform – enagage with the Pros to discuss the wine you want to taste/buy

Prioritized – while # 3 may give more engagement, it does not directly solve the user problem. The expert’s opinion may again may be subjective and the user may or may not agree to the recommendation. Also, since the user is buying from the grocery store, availability of expert to give real time opinion may not be feasible. #1 – is more applicable for search, may drive less enagement

However, # 2 really engages the user, by understanding what is the personality trait of the user, mood, and may help in providing better recommendations on the wines. Looking at the user journey, it keeps the user engaged at multiple levels

Features of Wine match predictor:

1. Small questionnaire about
   1. current mood
   2. personality traits
   3. Pallete choices
   4. what wine will they want to buy/drink today
   5. What does the general ensemble looks like – accompanying food, time of food (lunch/dinner/brunch) – have pictures of the food
   6. Who else will be drinking the wine – social connections
   7. Budget of buying wines, # of bottles per purchase
2. Predicts the wine based on data entered. and shows what kind of ensemble goes with it
3. Lists recommendations as per other users of the feature with similar tastes and moods ex: what wines do other similar users buy
4. comparison table between predicted and recommended wines
5. Comment, share, react to the recommended wines
6. Displays users selection (the wine that user wants to buy)
7. Post pictures of the wine – review and rate, share on timelines, share in other wine groups

Tradeoffs:

1. Privacy of the similar user – users may or may not be comfortable in sharing their recommended wines on social media, moods, tastes, ensembles, in which case, the recommendations may not be that accurate
2. Edge case where prediction is way different from recommendations given by other users
3. Accuracy of predictions
4. mood/food changes everyday, so the user may/may not like the wine the next day or in the coming days till the next purchase

Success metrics

1. # of users completing the questionnaire per session
2. # of predictions clicked per user per session
3. # of comparisons clicked per user per session
4. # of reactions/comments/shares per predicted and recommended wine per user per session
5. time spent per user per session
6. North Star – number of user selections displayed per user per session
7. Counter metrics – number of predictor wines selected per user per session