A strategy deck that lists "AI" and "Kubernetes" as strengths is not a map. Wardley Maps place user need at the top, a value chain of components below (what depends on what), and an evolution axis from left to right: genesis → custom → product → commodity. Components move right over time. The point of a platform bet is to notice you are building a custom thing that the market already treats as product or commodity, or the reverse: treating a genesis advantage as if you could buy it.
How to draw one in an afternoon
Anchor: who is the user (internal developer, end customer). Need: "ship a service with observability." Components: CI, clusters, golden images, secrets, the actual product code. Position CI: if GitHub Actions is fine, it is product/commodity. If you run a unique regulated pipeline, it may sit left — and that is a cost. Invisible components (power, AWS) sit near commodity.
Need: Deploy service
App (custom)
Golden path (custom → product if you productize it)
Kubernetes (product)
Compute (commodity)Climatic patterns: everything evolves; inertia (old contracts, skills) slows you. Doctrine: use common components where they are commodity.
Platform bets
Build a paved road when many stream teams duplicate custom work. Do not build a private cloud that clones EC2 unless you have a reason that survives the map (air gap, cost at insane scale). Buy observability if it is product; build the glue to your domain if that glue is still custom.
Maps are wrong; they are less wrong than a SWOT. Redraw quarterly. Fight about position (is our feature store genesis or product?) — that fight is the strategy conversation.
Pitfalls
Putting everything in the middle. Mapping org charts instead of components. Using the map as a wall poster nobody updates.
Read Wardley's introductory chapters (the book-on-medium). Then map "how a feature gets to production" with three engineers. If CI sits at genesis while the industry sits at commodity, you found a bet: migrate, or justify the custom with a constraint that is still true. The axis is the insight. The boxes without an axis are a whiteboard doodle.