It’s an uncomfortable number, but according to Pendo around 80% of features are barely or never used. They stay in the code, add to the user’s cognitive load and slow down shipping anything new. This is exactly where the idea of product garbage collection comes in. Garbage collection is a concept from computer science and it maps surprisingly well onto product development. It does exactly what it promises: it takes out the trash. ...
Paul Graham, Silicon Valley investor and an essayist worth reading, notes that there are two senses in which writing can be good: It can sound good, and the ideas can be right. The first one might be something you can actually calculate, so I gave it a try and vibecoded a small tool: https://readability-analyzer.vercel.app/ You paste in English or German text and get back a set of readability scores. No AI, no backend, nothing leaves the browser. ...
A remote control once just changed channels and volume. Today, when more than two thirds of Europeans use streaming services, the remote is no longer an accessory but the first touchpoint between the viewer and the product. And when that first touchpoint fails, it drags down the overall product experience. Learning from the Originals Zenith Space Commander 600 Early remotes like Zenith’s Space Commander from 1956 worked through simplicity. Each press made a click that confirmed the action. There was no latency and no guessing. ...
The challenge with AI isn’t just the technology. Often it’s how organizations are structured around it. Whether the work involves building internal AI infrastructure or shipping user-facing AI features, the ability to ship depends on how teams are set up and work together. That’s where Team Topologies comes in because if the org chart is a mess, no amount of machine learning will save it. The following four patterns outline how teams can be organized to support sustainable and effective use of AI. ...
It is 2025 and AI is well beyond the experimental phase. It is now embedded in the daily workflows of product teams. Language models assist with drafting OKRs, identifying gaps in strategy, screening interview responses, analyzing churn data and generating test scenarios from acceptance criteria. Most of these tasks are powered by large language models (LLMs). What Happens During Prompting When you type a prompt into ChatGPT or Claude, the model replies with something fluent and useful. But what actually happens inside? An LLM is a deep neural network trained on vast amounts of text. It generates responses by predicting one token at a time. This entire process runs on the Transformer architecture, the foundation behind models like GPT-4, Claude and Gemini. ...
With Team Topologies, Matthew Skelton and Manuel Pais aim to reduce the complexity of tech organizations to four team types and three modes of interaction. The end goal is to help teams produce software that aligns with customer needs and is easier to build, run, and own. (Skelton & Pais) Framework-Influences One of the central theses is that software architectures and the way teams interact are closely connected. This can be attributed to Conway’s Law, which states that an organization’s communication structures influence its software architecture. ...
What is the best way to split a cross-functional product development team? Here are six approaches with their respective pros and cons. 1. Feature Teams A cross-functional team with the end-to-end ownership of a feature. For example, an e-commerce checkout process or product recos. The members of the feature team work directly on all code bases required to further develop the feature, even if they are not responsible for that part of the software system themselves. ...
Product Repo is my personal, curated collection of product management articles that I find worth reading. The list is organized into categories such as product-led growth, strategy, leadership, product management fundamentals and many more. You can find it here: https://github.com/sischreiber/product-repo/ It’s essentially a GitHub readme file in my rather sparse repo, which I update sporadically whenever I come across noteworthy articles.
Glad to have my article on Mind the Product where I explain why we have a bias toward increasing the scope of our products, even though evidence suggests that subtracting features can increase the overall value for users, customers and the business. https://www.mindtheproduct.com/the-irony-of-product-led-growth/
Software engineers don’t go to product managers to understand how to build a software application. But when engineers start sharing their knowledge about product, there is often something in those words. Especially when the authors have left a footprint in the tech industry. Joel Spolsky on the Worst Possible Strategic Mistake Joel Spolsky is one of the co-founders of Stack Overflow. Although he has hardly posted on “Joel on Software” for years, there are several timeless articles worth reading. One of them is “Things You Should Never Do, Part I” from 2000. It is about a specific strategic mistake in the tech industry and why the cause lies in the rule of thumb that code is harder to read than to write. ...