Prioritizing Customer Feedback with Amazon Senior Product Manager

This week, Product School hosted Anu Venkatesh, Senior Product Manager at Amazon. Anu gives perspective on what top tech companies look for in a PM, and how Amazon keeps the product process focused on customer feedback and problems.

Meet Anu

Anu Venkatesh, Amazon Product Manager

Anu Venkatesh is a Senior Product Manager at Amazon Lab126, working on Alexa AI.

Before this, she was a Senior Speech Scientist at Nuance Communications working on their personal assistant technology and data export framework. Prior to that, she was a Senior Technical Consultant at Junction Solutions where she developed and executed a resource scheduler that was based on genetic algorithms.

How did you shift from being in a tech background to product?

Being in product is really about putting the customer first. We do it in all our roles. It just differs in terms of the degree to which we focus our energy on thinking about the customer problem versus the solution. For me, I started a company right out of my Master’s which really gave me my first tryst with Product and I loved it.

All my roles have been such that they involved customer interactions and that was on purpose for me to stay connected with the customer and build my product sense; and ultimately moved into a product role at Amazon.

Read next: Transitioning to Product Management From ANY Background

two computer screen on a black desk, backlit with blue light

What are steps you recommend to break into tech and get an interview with a top company?

Getting into any domain is a matter of curiosity. For tech especially, there are so many resources on the web today. Top companies are really looking for your experience and true product sense. From my personal experience, to get interviews with top companies, focus on demonstrating core product muscle of strategy, sense, prioritization, delivery and use of metrics to make informed decisions in terms of what you represent for your background on your resume.

How do you develop your own PM skills and learn new processes and templates? What resources do you find most valuable?

Thanks for this question. To me, curiosity is one of the key things about growing in any role. To keep developing my skills a few things I particularly do:

  1. Read blogs, articles, newsletters
  2. Connect with the community – Product School is a great one, and they have great events to learn from as well :). There are several such communities
  3. Build a network in your company. It is the easiest way to connect with more folks – learn about what they are working on, challenges they are facing and how they are thinking about solving them. Nothing like a real-world problem to brainstorm with colleagues and keep the juices flowing!

For processes and templates, I’m lucky in that Amazon is a massive company and has a huge set of resources I can lean on – I usually work with folks to identify the closest templates that match my needs to build off of them.

Check out our collection of templates! Product Management Templates

template on pieces of paper for the stages of setting up a business

What is more important during early stages of a product? Getting more users to try it and possibly generate revenue or building product with many features?

The earliest stages of a product, the first task should really be ensuring product-market fit. At Amazon, the working backwards process is where the process starts. This involves really identifying the customer problem and what it means to address it.

As part of this, you define the minimum set of features you would want to deliver. Identify what it means for your customers to use your product and what is the retention period. Is it something someone does every day (e.g. some social activities) or something that they do very infrequently (maybe even just once in their lifetime, e.g. purchasing a house). User studies are extremely useful in getting a signal around what customer needs and problems are. Leverage those mechanisms.

I also believe in failing fast. Identify the minimum product, put it out there with at least a few customers, get feedback and iterate quickly. That will help identify any gaps and validate your proposition with an opportunity to strengthen what customers are liking and fix what they don’t.

Could you share your thoughts on how success is measured for ML/AI based Products at a product and business level ?

robot hand touching white lights connecting in a network

Thanks for this question! Defining success criteria for your product is one of the most critical pieces of delivering a product. For ML/AI products, you also need to ensure you are tracking the quality of the backend models and systems e.g., precision, recall and accuracy. Plus there may be specific ways of measuring different tasks e.g. speech recognition, language generation etc., all have metrics that are commonly used.

It is important to set up an initial bar for the quality of these backend systems. Ultimately though, everything should be grounded in customer feedback. You may receive a low customer satisfaction score for your product even with a 90% accuracy on your model, which you may consider high; but now clearly based on customer feedback indicates is not enough. Business metrics are really what you believe is the vector against which your product is adding value to the business – could be users, revenue, retention, satisfaction ratings. I also encourage identifying mechanisms to identify how your product’s business metrics are impacting the broader business and corresponding lift attribution.

Read next: Tech Trends of 2021: The Rise of AI

Any final tips?

For any role you are in, try to get to the root of “why are we doing this/ what customer pain point are we solving.” Talk to your fellow PMs to understand why they chose to go down the path that they did. Establishing the core value proposition for the product and determining the persona and pain point to be solved is critical for a PM before you try to design a solution. Focus on building that product sense.

Enjoyed the article? You may like this too: