Frustration Architecture
Priority Churn
What matters keeps changing, so finished work is undone before it pays off. The team's labour evaporates without a funeral.
Dr. Tim Hough · ISBN 979-8-9965397-1-0 · Buy the book →
From the book
Priority Churn is the architecture in which the top three things every quarter are not the same top three from last quarter, and last quarter's top three were quietly retired without a conversation. The team built them. The team shipped them. Then the team was asked to be excited about something else.
The damage is not that priorities change — they should. The damage is that priorities change without acknowledgement of what the previous priority cost, and without a clear reason. Each switch teaches the team that finishing a thing is not the point. The point is to be in motion when the next priority is announced.
Priority Churn often gets praised as agility from above and experienced as whiplash from below. The architecture is the lack of an honest conversation about trade-offs — not the act of changing direction.
The diagnostic
If you can ask three people on the team what the top priority is and get three different answers — and they are all confident — you are inside Priority Churn. If you can ask what was the priority six months ago and nobody can remember the answer, the churn has already done its work.
The fix recipe
- 1
Force a stack-rank: write the top three priorities in order, with the trade-off each one is making against the next thing down.
- 2
Run a cadence audit — note every change of direction in the last two quarters and what the previous priority cost the team.
- 3
Before any new priority is announced, name what is being explicitly de-prioritised, and tell the team what work to stop.
- 4
Hold the rank for a fixed window. When it changes, walk the change through the Three Doors out loud — "I can't fix that right now. Here is why, here is what would have to change, and here is when we'll revisit it."
All five architectures
Most teams are living inside more than one at a time.
The diagnostic runs inside a workspace.
Public content ends here. The live tagging, clustering, and Three Doors decisions against each architecture happen once you stand up a workspace.
Source: Why Your Best People Stop Trying. Dr. Tim Hough.