The project ends well. There is a presentation, a document, a set of files, and a period of genuine enthusiasm in which everyone can describe what the company stands for. Eighteen months later a sales director is building a proposal deck in PowerPoint using a logo pulled from the website, the regional office has produced a flyer in a typeface nobody recognises, and the new SKU launched with a layout invented by the printer because the launch date moved forward two weeks.

Nothing dramatic happened. The identity simply had no owner, and an identity without an owner is a document rather than a system.

The failure is structural, not cultural. Consider who actually produces branded material in a mid-sized company: the sales team, daily, under time pressure, for a specific meeting. Regional agents, occasionally, in languages nobody at head office reads. Whoever handles the trade show, once a year, in the two weeks before it. HR, for recruitment. The plant, for labels and signage. Most of these people have never opened the guidelines, and none of them will consult a fifty-page PDF at seven in the evening with a client meeting in the morning.

They will use whatever is quickest to find. Drift is not disobedience; it is the predictable result of the correct option being slower than the wrong one.

Which suggests where to intervene. Not in more rules — the volume of rules is inversely related to how many get followed — but in three things.

A named owner. One person with the authority to say no, and the standing to say it to a commercial director. In most companies this authority is assumed rather than assigned, which means it is exercised inconsistently and collapses the first time a deadline is at stake. The owner does not need to be a designer. They need to be someone whose approval is required before something goes to print.

A short list of non-negotiables. Five or six things that cannot change under any circumstance, distinguished explicitly from the large number of things that can. Most guideline documents make no such distinction, treating the exclusion zone around the logo and the fundamental structure of the pack as equally binding. The result is that everything is technically mandatory and nothing is actually enforced. A short list of things that genuinely matter, stated in one page, is honoured far more often than a comprehensive document that is honoured nowhere.

And ready-made assets for the situations that repeat. Not templates in the abstract sense — the actual deck the sales team uses, in their software, with the slides they already build every week. The label template already set to the retailer's specification. The social formats. The recruitment ad. Every recurring need that has no ready asset is a future improvisation, and improvisations accumulate.

There is a fourth element, harder to install and worth more than the other three: knowing what the identity is for. A team that understands why the range is organised the way it is will make reasonable decisions in situations the guidelines never anticipated. A team that has only been given rules will either follow them mechanically into a situation where they do not apply, or abandon them entirely at the first difficulty. This is the argument for spending an hour with the sales team at handover — not walking through the document, but explaining the reasoning, so that the fifty cases nobody wrote down get handled sensibly.

The pattern to watch for is a simple one. Drift never announces itself. No single decision is large enough to raise, and each was made by someone acting reasonably under pressure. The company only notices when it assembles its own materials for a trade show or an investor and finds it cannot present a coherent picture of itself, at which point the conclusion drawn is usually that the identity needs redoing.

It rarely does. It needs someone whose job it is, a short list of things that actually matter, and the correct file being easier to reach than the wrong one. That combination costs a fraction of a redesign and is the difference between an identity that compounds and one that has to be bought again every four years.