Who decides
Who, in our organisation, holds the mandate to make these decisions? With the table of every decision on the site, who decides it and by when.
The question
Who, in our organisation, holds the mandate to make these decisions?
It is the question that makes everything else useful. A catalogue of decisions with no ownership produces awareness and no effect — which is exactly the condition this site starts from: the decisions exist, documented, but addressed to whoever executes them.
Three responsibilities the change requires and that rarely exist
Three responsibilities, and not for the same reason. AI visibility — a discipline the corpus places between Brand & PR (input) and Operations & Data (downstream traffic), measuring the output the model returns; it is anything but marginal in the corpus and is still missing. Trust calibration — five artefacts (the delegation scale, confirmation points on the irreversible, situated uncertainty signals, traceability of actions and sources, recovery paths) scattered across UX designers, content designers, PMs and legal; if nobody holds them together, they don’t get exercised — see Product & Design. Accessibility — a job no other function can take on and that the market does not cover; if it’s missing, the decision on accessibility has already been made: wait — see Product & Design.
This is not a call to hire three people. It is the observation that three decisions in this chapter have nobody preparing them.
The decisions with no natural role
A second group of decisions has no plausible home in any existing function, for a different reason: they are not product decisions.
- The moments when the relationship runs through us — a positioning choice disguised as a channel choice.
- Does engagement stay a KPI? — a decision about how the company measures the product, not about the product.
- The permission to contradict the user, inside Serendipity declines — a decision about what the company sells.
They belong wherever positioning gets decided. Who that is depends on the organisation: this card says the level is not the product’s, and stops there.
And there is a third case, the most common: decisions that have a natural role but no mandate to exercise it. The user’s language replaces the system’s taxonomy is a decision for the Information Architect about nomenclature that belongs to the catalogue, the ERP and the org chart. Without an explicit mandate it is not a decision: it is a research observation.
Every decision, in one table
The two tables are generated from the pages: if a decision changes owner or date, it changes here.
The cross-cutting decisions
The decisions inside the phenomena
| The decision | Who decides | By when |
|---|
An em dash in the “Who decides” column is not a missing data point: it is the content of the decision. Those are the rows to look at first, because they are the ones nobody is preparing.