Frame content-and-data programmes by change-cost tier
When proposing a content or data architecture programme, separate it into a backend-only tier (no ask of staff), a modest-asks tier (light behavioural change), and a full-reset tier (enforced behaviour change) — the tiers expose the trade-off and most engagements settle in the middle.
Content and data architecture programmes — wiki cleanups, document-store reorganisations, taxonomy work, knowledge-base consolidations — can be specified at very different costs to staff. A backend-only programme can do quite a lot without asking anyone to change a habit. A full-reset programme enforces structural change through policy. Most decision-makers want the upside of the latter at the cost of the former. Tiered framing is what makes the trade-off honest.
The proposal shape is three tiers, each cumulative on the one before:
Tier A — Backend only. Mechanical work that staff don’t have to participate in: retention labels, archival sweeps, bulk metadata application, search-noise removal, redirects. Lowest risk, lowest disruption, moderate uplift. The new structure becomes available; staff aren’t required to use it.
Tier B — Modest asks. Tier A plus a small set of behavioural expectations: stop saving organisational content to personal storage, follow a simple file-naming convention, use a single canonical location for each document class. Reinforced by light-touch correction (the named co-ordinator notices and flags regression) rather than enforced by policy. Higher uplift, durable if the co-ordinator has the remit.
Tier C — Full reset. Tier B plus enforcement: behavioural expectations encoded in policy, named accountability per functional area, a visible transition programme, training. Delivers the full target end-state. Most demanding, most durable in principle, most likely to fail in practice in organisations with a history of incomplete change programmes.
Two diagnostics emerge from putting the tiers side by side. First, the marginal cost between Tier B and Tier C is large: roughly three times the change cost for perhaps a third again of the uplift. Most engagements should land in Tier B unless there is a specific reason for the full reset. Second, the precondition for Tier C is sustained leadership backing for eighteen months or more — see Leadership team AI fluency must be collective, not individual — and the historical track record of that backing is the input that should drive the tier selection. Tier C is worth attempting only if you believe enforcement will hold; the history usually says it won’t.
The framing generalises to any programme that combines backend cleanup with behavioural change. State the tiers, name the staff-change asks at each level honestly, and let the decision-maker pick from a real set of choices rather than a single recommendation that conceals its assumptions.