How the drift happens
Nobody decides to have four versions of a logo. It happens through a series of individually reasonable decisions.
A dealer needs a board urgently and gets one made locally from a logo copied off a website. A regional office prints leaflets with a slightly different blue because the local printer built the colour from process inks. An old artwork file with the previous address stays in circulation because nobody knew there was a newer one.
Three years later there are four blues, two logo lock-ups and an address that is one office out of date, and no single person did anything wrong.
What it costs
Recognition, first. A brand that looks slightly different in every shop is a brand that is being recognised slightly less well everywhere.
Money, second. Every locally produced piece is bought at single-unit prices, and the aggregate across two hundred outlets is far more than centralised production would have cost.
And time. The recurring, unmeasured cost of deciding again which file is the right one.
The controls that actually work
One artwork master per item, held in one place. Not a folder of files - a named current version, dated.
A short, usable specification. Two pages that settle colour, type and lock-up, written for the person who will apply it rather than for a brand team.
Central production. The single most effective control, because it removes the mechanism by which drift happens. Outlets receive material rather than commissioning it.
Per-outlet kitting. Each dealer's set boxed and labelled, so what arrives is what was intended.
A route for urgent local needs. If there is no fast official way to get a board, dealers will make their own. That is a process failure, not a discipline failure.
Handling outlets that are genuinely different
Networks are not uniform: outlets differ in size, format, language and product range. Treating that as a series of exceptions is what makes central production feel impossible.
The answer is to design the variation in - a master with defined variable elements, versioned by size, language or range - rather than to make a new file each time.
Done properly, adding a hundred outlets is a data problem rather than a design project.
Fixing a network that has already drifted
Do not try to match the most recent version, which is usually the worst. Re-establish the correct one, then bring the recurring items back to it in a planned order.
Replace boards on a schedule rather than all at once, prioritising the outlets with the most traffic.
And close the mechanism - central production and a fast official route for urgent needs - or the same thing will happen again over the next three years.
