Customer-Driven Cross-Functional Teams with fmr Sr PM at Amazon

This week Product School hosted Karthik Sankar, former Senior Product Manager at Amazon for an #AskMeAnything session. Karthik discussed how to develop skills that make you a desirable Product Manager candidate. He also discussed how to form a customer-focused mindset when building solutions to problems and working with Engineers and Stakeholders.

Karthik Sankar speaker image

Meet Karthik Sankar

As Director of Product Management at Loblaw Digital, Karthik leads strategy and Product Management for the grocery delivery business of the largest retailer in Canada. At Amazon, Karthik was a Product Lead for Automated Brand Protection, where he envisioned, designed, and launched automated technology solutions to proactively prevent millions of infringements and counterfeits, using real-time catalog scanning, Machine Learning, and computer vision. 

Karthik also worked with the marketing team of Amazon Marketplace, where he launched influencer marketing and personalization initiatives by collaborating with successful entrepreneurs and business thought leaders.

What are the desirable qualities in a Product Manager and how should one go about developing the right skills?

In general, there are a bunch of things that interviewers look for in PMs. Here’s a quick list that hopefully helps as a guiding principle:

Are you able to structure your approach?

  • Start from understanding the customer and their pain points
  • Prioritize pain points
  • Identify solutions
  • Prioritize/trade-offs

Are you data-driven and goal-focused?

  • Are you setting the right goals and KPIs to evaluate success?

How are you handling ambiguity?

Do you have the skills to bring the product to life?

  • How will you build a roadmap?
  • How will you think about product design?
  • What is your approach to experimentation?
  • Are you thinking iteratively?


  • Stakeholder management
  • Negotiation skills
  • Handling disagreements and driving consensus
  • Handling failures & stressful situations

Do you have any advice for someone who has spent several years in software (as a Dev, SA, BA, PM) and wants to become a Product Manager?

For interview preparations, I would recommend two books: Cracking the PM Interview and Decode and Conquer. One of the best things about the Decode and Conquer book is that it presents a framework that every PM will apply in their daily life.

Another way to think like a PM and train our minds is to go back to a common question in Google PM interviews: “What is your favorite product, how will you improve it?”

working in an office

What is the difference in traits that you look for when hiring PMs in a large company like Amazon vs LobLaw Digital?

I don’t’ think there’s a difference in traits necessarily. In fact, take a look at my response to a previous question. The evaluation criteria, the responsibilities of a PM, and the approach to problem-solving. These are the same, be it a big organization or a small organization or a startup. One additional mental model that I carry with me all the time is Amazon’s Leadership Principles, and how they apply to a PM’s responsibility.

I use them to this day, every day, in how I approach problem-solving and how I make decisions, and how I hire, train, and build a strong PM team.

You may also like: 5 PM Skills for the 4th Industrial Revolution by pi-top Senior PM

How do you avoid sounding like a passion project when pitching the problem to interviewers?

It is important to spend a lot of time on the customer. Specifically, try to answer the following questions:

  1. Who is the customer?
  2. What is the customer problem or opportunity?
  3. What is the most important customer benefit?
  4. How do you know what customers want?
  5. What does customer experience look like?

Often, we are guided by our passion for a particular solution. While passion is great and very much necessary, it is important to go back to the basics and work backward from the customer. Be as data-driven and objective as possible. Often we do not know who the customer is.

Then, we falter in not understanding what their problems are, then we fail to understand what is the size of the problem, and if there are many problems, we will fail in prioritizing the most important ones. These are the key concepts we need to remember.

Karthik Sankar quote

What kind of metrics can you use to measure the success of an MVP? If I was able to get 30 signups for a portal in 6 days, does that mean my idea is good enough?

It is very important to set goals and measure against goals.

  1. Goals should reflect the problem you are trying to solve in the most accurate way possible. Keep them focused, specific, actionable, and with a timeline.
  2. Build a ramp to the goal. If it is a new product, and you are measuring sign-ups, start with 0 (before launch) and build a ramp – X in Week 1, X in Week 2, and so on.
  3. Goals should not be set in such a way that you can achieve it in your sleep, keep them challenging but realistic.

What are the big NOs that make you reject a candidate outright when hiring a Product Manager for your ogranization?

The biggest NO is the inability to think through any problem with a customer-focused mindset. If someone jumps to the solution, it’s a NO. As an interviewer, I do not look for the right or the best or the most creative solution, I am looking for the right approach to solving the problem for the customer.

How do you get approval and agreement from senior stakeholders when there is disagreement, especially when they aren’t technical or not as involved in the projects?

  1. Have a data-driven, clear, and specific approach to how you are presenting the customer problem: Why it is important to solve, what is the solution, why is it the right solution, and how will you implement it? Often, putting a lot of thought into this process automatically reduces disagreements.
  2. Understand the reasons for their disagreement. You need to spend a lot of time here, dive deep, to get to the root cause of the conflict.
  3. Acknowledge the concerns and value them.
  4. If the concerns are valid and a no-brainer, go back to the drawing board to fix them.
  5. If the disagreements are not straightforward, identify and mutually agree with stakeholders on the means to validate both viewpoints or one after another.
  6. Build 2-way doors, test fast, and iterate.
  7. Always build and track against goals – solving the customer problem should be the main focus of all stakeholders, keep an eye out on disagreements that are tangential to the problems you are trying to solve.

Is it possible for Product Managers to change industries while still remaining a Product Manager (Eg. Infrastructure to Software)? If yes, how does one go about doing so?

Yes, it is possible and happens all the time. PMs are generally valued for their ability to solve customer’s problems in a structured way, ability to handle ambiguity, and the ability to create impact. It is important to remember that the core transferrable skills are the same.

What you might need from an interview standpoint is a basic understanding of the product, and testing yourself on whether you are able to think about the customer of the product, what problems they might face, and how to solve them using the transferrable skills you have.

How do you deal with ambiguity?

The best way to deal with ambiguity is to split the problem into smaller chunks, keep an eye on the goals to prioritize the problems ruthlessly, and solve the problem iteratively.

This is easier said than done, it takes a lot of practice. Along the way, you should look at how you are progressing against your goals and measures of success. Also remember to keep building 2-way doors, through which you can walk back if the solution doesn’t work.

ambiguous windy road

Through your PM career, how did you handle the situation when feature request comes from execs in a top-down manner and they want the feature to be shipped ASAP and not giving you resources or time for customer research?

Great question. As PM, it is your responsibility to not accept feature requests at face value. Go back to the basics – who is the customer, what is the problem, is this the correct problem to solve (among hundreds of others), how to prioritize, how to estimate impact.

If a thorough analysis yields the insight that the feature should not be prioritized, it is important to present the thought process to execs, open up a discussion, and drive consensus. I’ve had to do this multiple times in my career – as long as the NO is supported by a strong WHY executives generally agree.

How do you deal with making a mistake? Especially when designers, engineers and marketers put in a lot of effort. How do you keep their trust and believe in decisions you make next time?

  1. It is important to understand what caused the mistake, and the size of the potential impact of the mistake. I do this by asking why.
    • Why did the mistake happen? Go one level deeper.
    • Why did that happen? Go one level deeper, and so on. Always look inward rather than blame others.
  2. Then, what will you do to avoid this and similar mistakes in the future – what processes, guardrails, or change the approach you will follow?
  3. Most importantly, include all stakeholders in this entire process of introspection – drive transparency and open communication as you work through the reasoning and collaborate with the stakeholders on the changes you will incorporate.

This is the key to earning trust, especially when you make a mistake.

people sitting in a room working together

How do you build rapport with your engineers when you are not as technical? Do you feel uncomfortable asking your engineers to do things all the time, and they have to put a lot of work?

One of the biggest mistakes PMs do is go to their tech teams and only tell them what to do. Do not do that. Involve the tech team early on, drive understanding of WHY you are solving a specific problem. Drive accountability and transparency of the business goals, collaborate with them all along the way.

Engineers are good at building things. They know that PMs are not technical. Engineers look to you to guide them on identifying the right problems to solve and prioritize. Listen to the engineers if they talk about technical constraints. Be ready to make trade-offs and deprioritization.

Does getting help from Subject Matter Experts or Tech Leads to better understand a proposed solution reduce the trust the tech team has in the product manager?

Not at all, seeking opinion from SMEs increases the trust in you as a PM. One key principle in earning trust is to listen for feedback, actively ask others to be the devil’s advocates for your thought processes. Doing this will increase the trust in your approach to problem-solving.

Say there is a mature product with high demand, usage and revenue streams, how would you specify a vision, and strategy to increase its uptake even further?

Go back to the customer and customer problems. What problems is the product solving, and what does it not. There would certainly be a lot of problems left to solve – start evaluating the size of those problems, prioritize, set higher goals, and implement them.

Think of products like Uber. every single time we use Uber, we think it is a mature product since it solves our problem. But, Uber keeps innovating to solve additional customer problems – which makes the product even more successful.

Any final advice for aspiring PMs?

For aspiring PMs, I urge you to spend a lot of time and effort understanding your customer and their problems. I see a lot of PMs fail in this step.

Did you miss this event? Check out our events page to sign up for the next #AskMeAnything session!

productcon banner

Enjoyed the article? You may like this too: