Example Google Interview Answers: Product Design - Google Camera?

As I was studying for the Google PM interview, I struggled with the lack of good, thorough example answers. You could find plenty of example questions online, and I even paid $80 for some practice materials that provided some high level frameworks. But even there, the answers were only "meh". I think the best way to learn is to deconstruct examples, see what other creative people tried, and take what works for you. 

So, I am going to take my notes for each example question I practiced, and share them. These are not the actual questions I was asked during the interview process, Google doesn't want you sharing those, which makes sense. Also, for what it's worth, I made it through the final round of interviews and to the hiring committee, but ultimately was not given an offer. These answers are the practice that got to me to the doorstep. I thought about not sharing them, but I truly believe shared experiences are the tide that raises all ships - as in makes us all smarter. So take from them, positive and negative, what you will :)

*************

Q: How would you design a Google Camera?

Background

The first thing I want to do is get more grounded as to the question that I am really answering. So some questions that I would want answers to if the interviewer could give them:

• Is Google Camera something that already exists? 
• Is it an addition to the Pixel phone, or is it a stand alone camera? 
• Is it for professional photographers, or photo enthusiast? Or is it for the every day photo taker? 

I will assume that Google is looking to make a stand alone camera, and that this is something that doesn't already exist. As for who the camera is for, I think any of the options I listed are on the table, and as we start talking about personas and use cases, we will be able to tease out which ones are most important.

Then, I want to know why Google would make a camera? It could be to differentiate the Pixel but I ruled that out. It could be feed more data into its software. It could be to diversify its business model portfolio, and have another hardware device differentiated by Google software that allows Google to sell at high margins to consumers.

I have a hard time imagining Google wanting to build the best physical lens in the business. That's not their specialty, and they would have the same suppliers as everyone else. So, my guess is that it has something to do with the intersection of Google software and a new stand alone device.

Goals and Constraints

I would say the goal of the Google Camera is to sell hardware differentiated by software, spreading the reach, ubiquity, and integration of its software while reaping the high margins that are possible in hardware.

The main constraints, then, are what software can be put into a camera with it’s very different interface. 

Personas

The way that I would think about personas in this case would be thinking what types of people can I think of that use cameras with high frequency?

I could have tried to build a persona around other potential camera problems, like time of day, photo editing, selfies, telescoping to space, etc. But we will roll with these three.

Requirements

Now I will try to write a story for a problem that we could solve for each of these personas:

A) As a college student who throws a lot of parties and events at my house, I want a camera that can take candid photos of our parties without me having to carry the phone around, because I want to enjoy the party without worrying about also capturing it. 

B) As a person in my 40’s, I love to go on hikes in nature and want a camera that can recognize the things I take pictures of, because I often don’t know what animals, plants, and landmarks I’m capturing. 

C) As a parent, I want a camera that can respond to voice commands, because it is always difficult to get pictures of the whole family when we go on vacation or on day trips together using the countdown features of classic phones.

We need to prioritize between these stories before we dive deeper and think of some solutions for one of these stories. Based on my goal and constraints, I would use a table like the following:

So, even though the "candid camera" story is the most difficult, that is the one I would try to build because I think it is a widely applicable use case, so a large target market. Users would be supremely delighted with such a product, and it would be a great learning opportunity for Google AI in terms of extracting data for computer vision.

Solutions

As far ways we could actually build a "candid camera" feature into a Google Camera, I can think of a few ways:

• All of them are going to be some variation of a tripod/camera integration. Let’s assume that the camera can turn itself on the tripod. A last assumption, is that you can set the camera into “candid party” mode. 
• One solution would build in software that recognizes smiles. The camera scans the area, wall to wall, looking for unobstructed shots of smiling faces, zooming in or out properly to get the “smiling” group in frame. 
• Another solution would be a camera that scans the room, taking photos on some regular cadence (every 30 seconds), and automatically deleting any obstructed photo.

Of those two, even though it would be much more difficult, I think the satisfaction of a camera that would trigger itself to take a photo when it saw smiling faces would be the most satisfying version of a candid camera feature of a Google Camera, and be a great way to improve Google’s computer vision technology at the same time. 

*************

Discussion:

One thing I thought to myself after finishing is that while I verbally feel good about the walk through I gave, it didn't lend itself to easy visuals. I might have tried to visualize more of it in hindsight. As usual, though, I'll leave it to you to deconstruct what lessons can be taken from this...

Other product design examples: Killer Feature, Maps

Other technical examples: Circle, Email Server, URL

Strategy example: Microsoft