ποΈ Move fast / PM Snacks 85
Move fast like Posthog or Lago β’ Powerful segmentation to increase impact β’ Focus on selling to escape the pain cave of pre-PMF β’ Use Blueprint prompts to accelerate
Hey everyone,
Sorry for the recent lack of newsletters, but I started to collaborate on an ambitious project that takes most of my time.
Donβt worry, I still found a cute place to work from, and spare time to share some new reads. Keep on reading π
Please support this newsletter by sharing it with colleagues or liking it (tap on the π).
Enjoy!
1οΈβ£ The magic of small engineering teams β’ Jul 2024 β’ 8 min read β’Β #hiring
Posthog shares how they operate and move so fast (their product really evolves at an impressive pace). In a nutshell:
1. Small, Empowered Teams: Small, autonomous teams make fast, effective decisions without heavy oversight, driving agility. Each team has one identified leader.
2. Direct User Engagement: Teams interact directly with users, ensuring product updates align closely with real needs.
3. Concrete Goals: Clear, deliverable-based goals keep teams focused on tangible value rather than abstract metrics like increase conversion by 10%.
2οΈβ£ How data sharpened the focus of product teams β’ Feb 2023 β’ 7 min read β’Β #data
Duolingo growth was stagnating in 2019, so the growth team went beyond the regular DAU metric, and created a model that segment active users in several categories:
By tracking the probabilities of users moving between different activity states, such as new users becoming current users or at-risk users reactivating, Duolingo could identify areas needing attention. The main lessons being: 1/ theyβre cracked at data-driven development, 2/ we need to break down broad user metrics into actionable segments, that we can experiment on.
3οΈβ£ 0-1 Path to Product-Market-Fit β’ Jul 2024 β’ 6 min read β’Β #execution
A long-time reader, Robin π, shared with me this slidedeck about ways to exit the βpain caveβ, and embrace a pragmatic approach to product development.
The most interesting advice is to sell to learn. This βeconomicβ interaction is the only way to reveal if the product genuinely meets customer needs, and gather accurate feedback. Rather than relying on broad customer segments, start by deeply understanding and working with a real customer. Sell once. And repeat.
π€ Stop Writing All Your AI Prompts from Scratch β’ Oct 2024 β’ 4min read β’Β #AI
The article suggests to use AI to write precise and targeted prompts. For activities you do regularly, you should create a βBlueprintβ prompt that hold a lot of the necessary knowledge that you have about the task. Take a peek at the example they wrote for educators. Then, try to create Blueprints for product management activities like: simulating a stakeholder meeting where everything goes wrong, preparing for a product reviews by brainstorming tough questions, and more.
β³ How we ship fast: our framework β’ Nov 2022 β’ 6 min readΒ β’ #execution
Most clicked link one year ago in PM Snacks 66.
In order to improve impact, Lago optimised their process to move fast. How do they manage to do it? They focus on a select set of topics and more importantly, they invest time beforehand to improve the quality:
1/ Spend time on the framing: define the usuals (problem, use cases, success measure, etc.) then go deep on the tech part (backend behaviors, graphQL features, design for each screen, etc.)
2/ Ask the engineers to dive-in and scope down accordingly
3/ Review everything: framing, scope, code and test a lot
β
Thank you for reading this far.
Until next time!
Olivier
Share β Please help me promote this newsletter by sharing it with colleagues (tap the refer button) or consider liking itΒ (tap on the π, it helps me a lot)!
About β Productverse is curated by Olivier CourtoisΒ (10y+ in product, startup co-founder,Β coach & advisor). Each issue features handpicked links to help you become a better product maker.