What we're learning, while we're learning it.
Product Craft: how we think and build. Product updates: what shipped.
Autonomous product intelligence
What it means for a product system to learn, adapt and act — without waiting to be asked. A complete guide to the emerging category.
Customer feedback management
A field guide to the discipline: staying customer-obsessed when there's more feedback than time to read it. Raise the floor, raise the ceiling.
The best AI product management tools in 2026
A practical guide to the tools turning customer feedback into shipped product — scored priorities, codebase-aware specs, and close-the-loop notifications. Here's how the categories compare, and where autonomous product intelligence fits.
How to write a product requirements document (PRD)
A PRD defines what you're building, who it's for, and how it should behave — the single source of truth that aligns product, design and engineering. Here's what goes in one, a template you can copy, and how PRDs are changing now that AI coding agents read them too.
All writing.
How to prioritise a product backlog without a meeting
The prioritisation process hasn't changed. What's changed is how long it takes. You no longer need a meeting to get to the answer.
Read →What good product specs look like now
The PRD format had everything except the thing that mattered. A good spec is a precisely framed human problem, with the context to act on it immediately.
Read →The first time you don't have to stop
Good enough and shipped always beat perfect and not live. That principle hasn't changed. What changed is where the ceiling is.
Read →Why solo founders ship the wrong things
The signal is usually there. The problem is what happens to it between arriving and becoming a decision.
Read →The bottleneck moved
AI coding tools changed the build economics. The constraint shifted from engineering to product decisions — and most teams haven't caught up yet.
Read →How to turn customer feedback into build-ready specs
AI coding tools are fast. The missing piece is knowing what to build — and having a spec ready when the AI asks.
Read →I spent nearly 20 years building spreadsheets. Then I automated them.
For two decades, I synthesised customer signal in spreadsheets because nothing else went far enough. Then AI made building cheap — and exposed the layer that had always been missing.
Read →The whole team, one direction
Projects, Ask annsa, multi-language support, a free plan. March was about making annsa work for teams — organised, unified and moving in the same direction.
Read →What product intelligence becomes
Three layers. Voice, behaviour, ambient. Layer 01 is live. Here's where product intelligence goes from here.
Read →AI product management tools in 2026: a comparison
Most tools use AI somewhere in the pipeline. Few cover the whole loop. A comparison of what's available, organised by where each tool fits.
Read →Using Cursor and Claude Code with a product management system
AI coding tools are fast. The quality of output is proportional to the quality of context. Here's how to connect customer feedback to your editor via MCP.
Read →How to close the loop on customer feedback
Somewhere in your backlog is a feature a customer asked for six months ago. You built it. They never found out. Here's how to fix that.
Read →