How would you solve Seattle’s traffic jam problem?

This is one of the problem solving questions asked by interviewers during Product manager interviews. Besides problem solving skills, product manager candidates are tested on product design, product strategy, market size estimation, and technical capability categories.

The reason why they ask problem solving question is that product managers actually face with this kind of big and ambigous problems on the job. But most of the time, resources they have are limited to solve these problems. Resources are time, budget, and engineering or design talents etc..

Interviewer doesn’t know the exact answer, because often times there is none. What they look for is how you take this ambigous and big problem, breakdown into manageable pieces, prioritize them, pick the ones and come up with ideas and solutions for those pieces to have the biggest impact in solving the actual big problem.

Let’s get back to our question:

“How would you solve Seattle’s traffic jam problem?”.

Of course, you have to first understand the problem by asking bunch of follow up questions like

What do you mean by solving the traffic jam? (Objective)
Is it reducing the traffic jam by X%? (Success Metric)
Is there any other factors I should be aware of? (Limitations)?
So on…

Assuming nothing is odd (like reduce traffic by 75% etc..), we can start wrapping our head around the actual problem.

The very first step I would take is to analyze the actual problem into pieces. Traffic jam is a result of some root problems. It is causation. Lets see what those might be:

  1. Lack of infrastructure — Roads and bridges are just not enough to handle number of cars on the road.
  2. Density of businesses — Most of the businesses are located in downtown so employees and customers need to commute to the same area. That is sometimes a source of citywide traffic congestion.
  3. Road Conditions — Because roads are not great, and traffic lights and signs are not smart enough, drivers need to slow down to achieve better control. This causes every other cars behind it to slow down as well.
  4. Weather Conditions— In city like Seattle where more than 200 days of a year are rainy, weather condition is the major cause of traffic jam as drivers need to drive more carefully.
  5. Cars — Think of a car failing in the middle of the road. It could be an engine failure or anything that stops car to operate. Lane is blocked until towing truck arrives.
  6. Drivers — Humans are not good drivers at all. They drive horribly, cause accidents regularly, and even some of them sleep while driving.

Yes! car accidents. But that is result of one or more of above causes.

Now that we know what the causes are, we can figure out which specific causes we should focus on to eliminate/reduce so we can have the biggest impact on traffic jam reduction. I.e, prioritization.

Before we eliminate any of the causes, we can brainstorm and come up with ideas to solve each piece. Here is the path I would follow:

  1. Lack of infrastructure —Either reduce number of cars or expand infrastructure. For example: build underground transportation, 
    Build another layer of road on the air, introduce new vehicle that doesn’t use road, or make flying car etc..
  2. Density of businesses — Either move some business to new areas or reduce number of cars commuting to these business. For example: Incentivize business to relocate, encourage employee bus program, Introduce work from home a day per week (20% reduction), shift working hours of some businesses etc..
  3. Road Conditions — Solution here is obvious. Detect bad roads and fix them. Get drivers use smarter mapping tools for navigation. Make traffic lights smarter with sensors etc..
  4. Weather Conditions — Either make cars better to deal with weather conditions or train drivers so they can drive better in bad weather. For example: Mandate tires to be seasonal, check lighting, require rainy day in driving school/test etc..
  5. Cars — Thorough check up of cars before they roll on the road. If not passed, get rid of those cars. For example: in-depth check up during license plate renewal, incentivize people to buy newer car, have emergency sideways and facility along the road…
  6. Drivers — Eliminate or reduce human interaction on driving. For example: Sensor technology to help drivers to react, self-driving cars

From there, discussion can go into any direction based on constraints and success metrics we might have. If interviewer wants you to choose the most effective solution that will reduce traffic by largest percentage by ignoring time and cost of implementation. Or maybe to come up with “quick win solution” that would require least amount of time and cost.

Lets say, we decided to come up with most effective solution. In this case, I would choose the cause “Drivers”. Because I know from my own experience. I commute everyday to my work. Although sometimes I am stuck in traffic due to weather conditions, It is “accidents” for the most of the time. I suspect that those accidents are mostly due to drivers. It is an hypothesis, and can be confirmed (or negated) by looking at historical records on past accidents. That is an different issue.

So, what if we eliminate human interaction in driving at all? Wouldn’t it be traffic jam free world where all cars are self-autonomous and are conscious of everything in and around itself?

The question has now become:

How do you build self-driving car?

which is a product design question. I will try to cover product design cases in my future posts.