PM Life: Art of Feedback (😁vs 😒)

Rose Yao
4 min readNov 18, 2018


I remember being an APM at Google, sitting in my first product review and presenting my first PRD (product requirements doc). I was so nervous. I don’t remember the exact feedback but I remember it didn’t go well… Luckily, I had great partners on the design and engineering side who helped me get better at thinking through product requirements and mentors who taught me how to do a better job communicating the why. Throughout the years, I’ve gone from presenting the PRDs to mostly reviewing the PRDs and I’ve noticed that feedback fall into two main types.

Type 1: Critical and Specific

This type of feedback that is very focused and generally leads with opposition. It assumes a right/wrong. Some examples we’ve probably all dealt with:

  • I don’t think a radio button is appropriate for this interaction.
  • I don’t like this element of the design, the colors don’t represent our brand.
  • You are designing for the wrong use case.
  • This is too complicated/impossible/not important.

Type 2: Curious and Exploratory

This type of feedback lead with questions and fewer assumptions. Some examples:

  • Who do we think the target users are? What are their goals?
  • What if <fill in the blank> was true? Would that change anything?
  • What if we removed this constraint?
  • What does success look like? What does failure look like?

Is Type 1 bad and Type 2 good?

Sadly, PM Life is rarely that black and white.

Type 1 is an easy fallback. It’s easy to find what’s wrong with an idea. It can feel good to give concrete feedback. But often it is frustrating for the team and limiting. However, there’s a time and place for Type 1 feedback. For example, when it’s time to fine tune a product or design, it’s important to get into the details and be specific. It sends a message that details and quality matters and we are going to sweat the small stuff. As a PM try to tie that feedback back to a product requirement that everyone agrees on. Leave the details of the design and technical implementation to your senior design and engineering partners. Examples of good Type 1 feedback:

  • Given the primary audience for this product is in Germany, we should take a look at the design in that language to make sure the strings fit.
  • We’ve seen in user studies that people are more patience when there’s a fun animation to help with this long initial download. Can we make that happen for launch?
  • We only get one chance to set an impression with this new product. Let’s take another look at what we de-prioritized and see if there are quick win that help with the first impression.

Over my career, I’ve started leaning more and more on Type 2 feedback for two main reasons: Ideas are fragile and I’ve stopped believing I know everything. It’s often too easy to lead with assumptions and critical feedback (Type 1 feedback), but the best ideas often seem a little silly or crazy at first. Take Snapchat with its disappearing messages and selfie filters. Or how many years AI has been a purely academic idea requiring super computers and then then suddenly it runs on mobile devices in real time with 90%+ accuracy. It isn’t easy to check our assumptions at the door and keep an open mind, I don’t always get it right. But here are things I try to do when I am reviewing product pitches and PRDs.

  • Lead with questions and my curiosity. First round of feedback, try to focus on questions vs answers. Understand the assumptions behind the idea.
  • Understand the work that’s been done. What are the user studies, data, feedback or intuition this idea is based on.
  • Understand the intentions, what are the goals of this product?

Usually once we agree on the assumption, the work, and the intentions, specific feedback is almost unnecessary. The team will naturally agree on a decision or understand what changes are needed to refine the product.

We are all in the midst of thinking through what 2019 looks like and the dreamers are pitching crazy new ideas. I hope this reminds us to keep an open mind and ask lots of questions before saying no.


Btw, we will still say no after asking questions, but link the no to a good reason so people aren’t afraid to come back with new ideas. Example: this is a really interesting new market you’ve identified but we need to focus on winning our core customers before investing in a new segment. Let’s revisit in ~6 months.


I do hope we all get to say yes to something new and a little crazy. That’s how the world gets better or at least more interesting.



Rose Yao

I spent the last 16+ years building products mostly at FB and Google. Also a food, travel, and fitness addict. Follow me @dozenrose or on