6 Steps to Turn a Mediocre MVP into a Quality Product
6 Steps to Turn a Mediocre MVP into a Quality Product
Proven tactics to uplevel quality with real examples from the trenches
Dear subscribers,
Today, I want share 6 steps that I take to turn a mediocre MVP into a quality product.
It’s easy to say “we care about quality,” but it’s another thing to make the daily decisions and sacrifices needed to deliver something that you’re truly proud of.
Let’s discuss the steps below with some real examples from the trenches:
Only the paranoid survive
Don’t go chasing waterfalls
Test your own shit
Ship to concentric circles
Marketing is the product
Launch is just the beginning
This post is brought to you by…Satish
Satish (ex-Meta product leader) and I agree: “The best way to scale yourself as a product leader is to empower everyone on your team to be mini-PMs.” In the free lesson below, Satish shares his playbook on how to scale yourself and your team better. He’s a dear friend and a top rated PM instructor — check out the lesson now:
1. Only the paranoid survive
This first step is more of a mindset. To set the bar on quality, you have to:
Assume that anything that can go wrong will go wrong.
Being paranoid isn’t fun, but it helps me:
Put myself in my users’ shoes
Consider edge cases
Sweat every detail
For example, when I worked on Roblox’s AI Assistant, we went out of our way to break it with adversarial prompts. We uncovered many bad answers and hallucinations that helped us improve the Assistant’s responses before launch.
2. Don’t go chasing waterfalls
It's time to retire the classic "waterfall" product development process. In fact:
If you’re handing a “complete” PRD to your designer, then you’ve already lost.
The best designers want to contribute as early as possible. So instead of a full PRD, I now share a one-pager with the problem, goal, and high-level requirements to kick-off the PM-design collaboration as soon as possible.
When iterating on designs, I like to:
Validate with real users early and often. Don’t wait for a design crit or UXR study to get feedback. Share in-progress designs in your user community weekly.
Think beyond the happy path. Details like product copy, empty states, and edge cases all matter. Obsess about getting them right with your designer.
When in doubt, narrow the scope. It's better to solve a single user problem exceptionally well than to tackle many problems poorly.
For example, I’ve been working on a product with complex workflows. My designer and I were nevertheless able to simplify the user experience by iterating with real customers and relentlessly focusing on details like product copy. Always take on the burden of making things obvious instead of putting it on the user.
3. Test your own shit
I recently came across a Reddit thread asking: “PMs, do you test your features yourselves?” The answers are why people hate on product managers:
As a PM, you have to find time to test your own shit.
You can’t outsource testing to your engineers or QA team because nobody has as much context on the product as you do. So please:
Make time to test. Stop polishing internal artifacts and instead focus on improving the product that customers actually use.
Lead by example. I work with a technical director who’s always the first to thoroughly test new products. Nobody is too senior to dogfood.
Test like a new user. Go through your flows as if you're using the product for the first time. Don’t assume that users will just understand how it works.
For example, at Stripe, everyone is encouraged to write friction logs detailing all the points of friction they experience when using a product. Here’s an example of a friction log (along with my template for you to make your own):