What the Work Produces
The artefacts, and the decisions they exist to improve.
Every organisation owns at least one. The handsome document that cost real money, filled a room with nodding, and then changed nothing at all. It is still there, in a drive somewhere, still handsome. Open it and the thinking inside holds up. That was never the problem. The problem was the job it was built to do, which was to be finished rather than to be used.
These are different briefs, and they produce different objects. A thing built to be finished is optimised for the moment it is handed over: complete, polished, closed. A thing built to be used is optimised for every moment after that, when a tired team reaches for it while deciding something real.
An artefact is a decision instrument
Here is the reframe that changes what you make. An artefact is not a deliverable. It is a decision instrument. Its purpose is to help a specific group of people make a specific choice better than they would have made it otherwise. Once you hold that idea, a demanding little question attaches itself to everything you produce. What decision does this improve? If an output cannot answer that, it is not finished. It is merely formatted.
If an artefact can't name the decision it improves, it isn't finished. It's formatted.
The polish paradox
There is a strange trap in finishing things too well. A highly polished artefact sends a quiet signal: the thinking is done, the editing is over, this is complete. And a complete thing invites admiration rather than participation. People receive it, appreciate it, and file it. Nobody picks up a pen to a masterpiece.
A working document with a few visible rough edges does the opposite. It signals that the thinking is still live, that a reader's mark would be welcome, that this is a surface to work on rather than a monument to walk around. The rough edge is not sloppiness. It is an invitation, and invitations get accepted.
Four jobs
The full list of things a behavioural practice can produce is long enough to overwhelm. Presented as four jobs, it becomes a conversation about what is actually needed.
- Understand
- Where behaviour stands now, and why. The behaviour map, the friction map, the diagnosis, the risk register. These improve one decision: is this worth acting on, and where?
- Decide
- What to change first, and what to leave alone. Prioritised shifts, decision principles, the strategy on a page. These improve: where does effort go first?
- Build
- The change itself, shaped for real conditions. Intervention concepts, prototype tests, the adoption pathway. These improve: what exactly do we change?
- Hold
- How the change survives contact with time. The measurement framework, the team playbook, the toolkit, the operating model. These improve: how do we keep this without the expert in the room?
Get the full guide
The complete guide, with worked examples, the running order for a team session, and all the detail the excerpt leaves out, is available as a free PDF.
Free. No spam. One email with your download link.