Nobody notices drift while it is happening. That is its defining property. Every individual departure is small, justified and approved by someone reasonable, and the aggregate is only visible when you put eighteen months of output on a wall at the same time.
Which, incidentally, is worth doing once a year. Print everything. Every market, every channel, every deck. Stand back. The room will tell you more than any audit.
Here is the typical sequence.
Month one. The platform lands. Everyone is aligned, because everyone was just in the room where it was explained. This is the high point and it is misleading, because alignment created by a presentation has a half-life of about a quarter.
Month three. The first legitimate exception. A campaign needs a different tone for a specific reason, and the reason is good. It is approved as a one-off. It is now a precedent and nobody has recorded that it was meant to be temporary.
Month five. A new agency is appointed for a specific channel. They are briefed on the platform, they interpret it, and their interpretation is competent and slightly different. Not wrong. Different. There is now a second version in circulation and no mechanism that notices.
Month eight. Someone senior joins from a business with a strong brand culture and, entirely reasonably, brings their instincts with them. Their instincts are good. They are also from somewhere else.
Month eleven. A second market localises properly, which is correct, and in the absence of a clear rule about what is fixed and what is local, they localise more than the centre intended. The work is good. It is also further away.
Month fourteen. A performance team, measured on conversion, discovers that a blunter, more generic message tests better in the short run. It does. Short-term response almost always favours the generic, which is one of the quiet structural forces eroding brands everywhere.
Month eighteen. Someone asks why the brand feels inconsistent. An audit is commissioned. The audit finds four tones, three colour treatments and two distinct value propositions. Everyone is surprised, and nobody can point to a single decision that caused it, because there was not one.
Then, usually, the wrong conclusion is drawn: the platform was not strong enough, so we need a new one. And a new one is commissioned, launched at an event, and the sequence starts again on a fresh eighteen-month cycle. I have seen businesses do this three times in a decade, which is an expensive way of never having a brand.
What prevents it is not more rules. It is ownership and a rhythm. One named person whose job includes coherence. A quarterly review that looks across markets and channels rather than within them. An explicit, written distinction between what is fixed and what is local. And a mechanism for approving exceptions that records them as exceptions with an end date.
Unglamorous, all of it. But drift is not a creative failure, it is a governance one.