Best Of Product Roundup - May 4, 2024
The best expert-curated product content from the best product thinkers from around the Internet delivered twice a month.
Every two weeks we consume all the product content on the Internet (free and paid), analyze it, and then we round-up links and short excerpts from the best content to:
Save you thousands of dollars per month in unnecessary subscriptions
Help you avoid spending days per week just reading instead of doing
And most importantly, we save you years spent trying the wrong advice.
Finally, we promote thinkers in product who are doing great work whether well known or soon-to-be-known!
Top Long-Form Articles
What we liked: A quick to understand critique of what’s wrong with most roadmaps from a brilliant mind in the field.
Choice Quote: “If you decide to make parts of it “public,” then do so only in the very short term and only for select things. Customers do care about features as a language and understand them so that lines up quite nicely. Resist the urge to go beyond a few months at any cost, though.”
How Reddit grew to 1 billion monthly users
What we liked: A thoughtful and yet not overly detailed analysis of some of the key phases and drivers in Reddit’s growth (the consummate community product) + a nice set of a lessons for Product at the end.
Choice Quote: “This also meant that communities were being started on intent-basis (you create a community of interest and attract users globally) vs. geographically bound communities that were prominent on reddit in the past. The new phenomenon of “intention-based community” also helped Reddit since in modern times and specifically even more so during COVID-19 it became hard to form local communities.”
What we liked: Finally an attempt to break the death grip of outdated “canvas” styles for stress testing ideas! The author creates a new “startup canvas” that includes just enough boxes to be useful with none of the extraneous content.
Choice Quote: “The Startup Canvas combines my refreshed Product Strategy Canvas with a business model. And it’s designed specifically for new products.”
What we liked: A concise explanation of how engineering can tackle the challenge of LLM non-determinism without expending too many cycles and still having a reasonably consistent product.
Choice Quote: “Unlike traditional software, where you can define a set of fixed inputs and expected outputs, LLMs are inherently non-deterministic. Feed the same input to an LLM multiple times, and you might get different outputs each time. This non-determinism makes traditional unit testing approaches ineffective for LLMs.”
Top Tweets
Keep reading with a 7-day free trial
Subscribe to Emergent - Product Newsletter and Podcast to keep reading this post and get 7 days of free access to the full post archives.