👨🔬 Using data from experiment to launch / PM Snacks #54
Learn to be a better leader by being demanding & supportive, how to use data to derisk, improve and launch new features, and more...
1️⃣ Demanding & Supportive
Jan 2022 • 3 min read • #manageyourself #leadership
Ravi Gupta shares some behind-the-scene stories at Instacart to illustrate how the best leaders can be demanding & supportive. It definitely struck a cord with me since I often made the mistake of being too nice and not pushing my team enough…
Most people think of demanding and supportive as opposite ends of a spectrum. You can either be tough or you can be nice. But the best leaders don’t choose. They are both highly demanding and highly supportive. They push you to new heights and they also have your back.
2️⃣ From experiment to launch: how data shaped a new comments experience
Jan 2022 • 5 min read • #data #execution
You may have noticed that comments finally became very visible in Figma. This article tells the story of how it came to be (and it’s a masterclass):
1/ They started with their vision: Figma is all about connecting everyone to the design process and improving collaboration. Yet, comments, the most collaborative feature was dramatically underused. Before moving one, they used data to prove that this vision led to more retention & plans upgrades (= the outcome).
2/ They framed the problems: through some user research & experts’ reviews they decided to work on this feature’s visibility. But before moving on, they implemented and then measured the simplest solution possible: prompting all developers to add comments. Spoiler: visibility was indeed a big problem.
3/ They delivered small increments regularly: after designing an entirely new experience they decided to break it into small parts that could be delivered and measured regularly. Thus lowering risks, and most importantly, acknowledging that tools used on a daily basis should change slowly to keep users productive. What an improvement since the big rollouts of facebook redesigns generating protests (when people still cared about it 😅).
3️⃣ How to make an analytics tracking plan, and get your team on the same (Notion) page
Apr 2020 • 6 min read • #data
Lucky enough to be a data informed team? Then you are probably maintaining a tracking plan somewhere… At uku, we are using Notion for this too. And we followed most of the advice shared in this article:
1/ Start with questions: don’t try to measure everything or you’ll just create noise. Instead, list the questions you want to answer to make better informed decisions.
2/ List the minimum required events/metadata: document these events, theirs matching metadata & implementation statuses in a Notion table.
3/ Don’t stop there, decide on leading and exhaustive metrics: try as much as possible to create metrics that encapsulate a success criteria. Imagining that you are a marketplace, you could consider your customers activated when they viewed +1 sellers for example.
4️⃣ Why Product Managers Don’t Need Domain Knowledge to be Effective
Jan 2022 • 7 min read • #hiring #team
This is an interesting take that I’ve seen realised numerous times in my previous experiences. Given a choice between a great PM without domain knowledge & an average PM with domain knowledge, you should always pick the person with great PM Skills. Unless you are in a very specific niche (with lot of regulations for example), teaching people product craft is usually much more difficult.
More importantly with expertise some biases can arise like the illusion of knowing better, inability to think laterally or simply a lack of creativity in problem solving. This is why when you join a new company you are usually asked to write an “astonishment report”, only a beginner’s mind can notice some things.
A warm welcome to the 12 new subscribers, we are now 1909 readers strong!
Thank you for reading me.
Have a great week!