PM Wargame: Design an app to help people pick wine at the grocery store

The board

You are tasked with designing an app that helps customers pick wine at the grocery store. 

You are starting from scratch; there is no existing app or functionality to work from.

---

In ‘PM Wargames’, I take on a hypothetical PM challenge and attempt to unpack the strategic & tactical approaches I would take to solve it. Unless sources are specifically cited, all information (including the challenge itself) is purely for illustrative purposes.

Game plan

There's a lot of information on the board, already.

The prompt provides us with a lot of context and expectations:

1. Customers will use this app while in a grocery store. This tells me a lot about the type of experience we need to deliver. It needs to be speedy. It needs to be convenient. It needs to lead to an immediate, meaningful, and actionable outcome.

2. The expected action is clear: help customers pick a wine. Clear job-to-be-done (JTBD)! It's about picking the wine; not learning about wine, not evaluating wines, not reviewing wines, not tasting wines, etc.

We’re in luck. Our goal is narrow and well-defined. That’s great.

We can focus on navigating the user from A (stuck at the grocery store) to B (picking the wine) as directly as possible, and disregard any fluff.

Identifying our customer

Based on the above, we can start identifying our customer profile and assign them a few attributes:

1. They're in a hurry. Standing in a shop likely with a trolley full of groceries, some of it defrosting.

2. They have purchasing power. Money to spend on wine.

3. They're not wine experts. If they were, they wouldn’t need an app to help them pick.

4. They're likely looking to impress. If they weren’t, they’d pick anything with a fancy-looking label. Instead, they’re doing a bit of research. Perhaps going on a date, having friends over, etc.

5. They care about taste. Enough to want a wine that works with dinner, not enough to have planned the wine before going to the shop.

Drafting an MVP

Let’s brainstorm some features for a ‘wine app’:

• Integration to grocery stores (stock, SKUs, etc.)

• Requirement gathering flow

• Recommendation engine

• Prices, deals, coupons

• Tasting descriptions

• Ability to purchase wine

• User-sourced wine scores

• Social features (social share, in-app profiles, etc.)

• User forum (e.g. to ask for recommendations)

• Peer reviews

• Wine pairing recommendations

• User profile (budget, stores they attend, location, age, preferences, etc.)

Let’s stop here. That’s a good list of ideas, some greater than others. Where do we start?

Thankfully, we’ve already done the work of narrowing down our wargame into a JTBD and a rough customer profile. Hopefully, picking the right ideas should now become a breeze…

Action

I want to design an extremely simple app. 

First, I want to gather a few yet very specific characteristics about my users: their rough budget (per bottle) and the shop(s) they frequent.

This will keep recommendations relevant and actionable (we want to avoid the frustration of not finding the recommended bottle).

Second, I want to emphasise with my customer. Here’s their likely situation:

They’re in a hurry. They're currently standing in the alcohol aisle, staring at rows of wine bottles, a trolley full of groceries (some of it defrosting), and needing to make a decision right now. Their friends are coming over tonight. The meal had been planned days ago but the wine was left to the last minute, again.

What should they do?

Open our app.

I want the customer to immediately be thrown into a requirement-gathering flow. The customer answers a few, plain-English questions (no jargon) about their situation:

• Are you serving meat? Which one?

• Are you having dessert? Entrées?

• What time of day will you be drinking the wine?

• Is it warm or cold out today?

I will plug their answers into an internal search engine.

Since I've been given this task, I assume we at least have a wine database of some sort. If we don’t, we can explore APIs like Wine Vybe or Wine Searcher.

We could simply return a list of bottles. Fine.

But I want to do better.

Wine can be a tricky purchase. My customer likely wants to impress. I want to reassure them that they're making the right choice by picking this bottle.

To do so, I will use generative AI to create a short bit of text describing the wine and tying its characteristics to my customer’s requirements.

Finally, I want to make sure this entire experience is quick and painless — getting them from panic to recommendation within a minute or so.

This is my game plan. What do you think of this approach? How would you play this PM wargame?