Main idea

A glossary, a one-pager, a PR/FAQ, a hierarchy of metrics, a decision log, escalation rules: the toolkit is reassuring, and its existence is easy to mistake for the existence of the alignment it is supposed to produce. The only criterion separating a useful tool from its imitation is what it compels someone to write down in black and white.

Each of these objects has its empty version. The decorative glossary lists words that posed no problem. The one-pager fills up with sentences broad enough to suit marketing and product alike. The PR/FAQ becomes an exercise in internal communication. Metrics pile up without anything saying which one comes first. The decision log records choices without ever naming what was sacrificed. The escalation rules exist in a document nobody activates. In each case the form is intact and the constraint has disappeared.

The effect is then not neutral, it is negative. A tool that lets you avoid a difficult conversation adds documentation to an organization that still has not decided, and it turns a disagreement into a polite formulation. So there is a single question to put to each object: what does it force you to make explicit? If it forces nothing, it aligns nothing.

Why it matters

This gives an audit criterion you can apply to an existing setup, without having to judge its writing quality: take each governance artifact and look for the thing it was impossible not to write.

It also strips authority from the equipment argument — "we have a glossary, we have a decision log" — which proves nothing about the state of the arbitrations.

Nuances and limits

Forcing is not always desirable at the same moment: on a subject still being explored, a tool that demands you settle the target or the main criterion manufactures a premature decision.

And a tool can genuinely constrain and still be useless if what it makes explicit interests nobody: constraint is a necessary condition, not a guarantee of relevance.

Open questions

• How do you recognize, reading a one-pager or a report already filled in, the difference between a cautious formulation and one that dodged the constraint?