marketplace incentives
Users you want the most need you the least
The best users in any two sided marketplace will already have a system that works for them.
Read noteNotes
Patterns worth paying attention to
Short notes on AI, systems, incentives, product, and the messy ways people and technology interact. A running archive of things I keep noticing.
product distribution
Distribution matters more now that building is easiy
As AI lowers the cost of building, the harder problem shifts to attention, trust and distribution.
Read note AI Product
The handoff is the product
As AI moves into the core of products, designing handoffs between human judgment and machine output becomes a central problem.
Read note metrics incentives
Track the metric. Remember the goal.
Metrics are only proxies. Necessary, but not sufficient.
Read note marketplaces product
The quality of supply determines the quality of demand
In most marketplaces, the kind of demand you attract is shaped by the quality of supply you build first.
Read note adoption user-behaviour
Your product is not the user’s goal
Users do not wake up wanting to use your product. They wake up wanting to get something done.
Read note funnels problem-solving
When in doubt, check the funnel
Funnels are one of the simplest ways to break down messy product problems into solvable parts.
Read note user-behaviour growth
Before you enforce rules, understand the resistance
When users do not follow your product rules, the instinct is to penalise. Sometimes the better move is to understand why.
Read note scope user-behaviour
Your product does not exist in isolation
Forces outside your scope affects product metrics. And it is still your job to understand them.
Read note community product
Find something you care about. Build from there.
One of the oldest playbooks: start with genuine interest, build a community, and let products emerge from trust.
Read note marketplaces supply-demand
A marketplace is only as good as its best participants
Supply and demand existing is not enough. A marketplace only works if the best participants find enough value to stay.
Read note friction behaviour
Defaults are product decisions
Most users do not actively choose. They accept what is presented to them.
Read note product systems
Every workflow has an unofficial version
The documented workflow is rarely the real workflow. Pay attention to what happens outside the system.
Read note product metrics
A product metric should explain a decision
If a metric does not help you decide what to build, fix, or investigate, it is probably noise.
Read note product analytics
Data tells you what happened. Users tell you why.
Do not hesitate to speak to your users. They can help more than you think.
Read note Product Thinking User Intent
The most useful product sentence I know
Before you debate features, think about this: what is the user actually trying to get done?
Read note