A company reaches a point, usually somewhere between the seed round and the first real marketing hire, where the material stops feeling like it belongs together. Someone says the brand looks tired. Somebody else agrees. Within a month there is a line item for a rebrand.
Sometimes that is right. More often the brand is not tired at all. It is fragmented, which looks similar from the inside and has an entirely different cause, a different cost and a different fix.
What fragmentation actually looks like
Fragmentation happens when the number of people producing material exceeds the number of decisions that have been written down. It is arithmetic, not carelessness. Every time somebody makes a slide and the type size is not specified anywhere, they choose one. Every time somebody needs a second colour for a chart, they pick one. None of those choices is wrong on its own. Fifty of them, made independently over a year, produce a company that appears to have no opinions.
The symptoms are consistent across every team we audit. There are three or four versions of the logo in circulation and no agreement on which is current. The deck has two different blues, one of which came from a screenshot someone colour-picked. Headings are set at four sizes that are close enough to look like mistakes rather than a hierarchy. The website footer and the invoice use different spellings of the company's own tagline.
Individually these are invisible. Cumulatively they read, to a customer, as a company that does not check things. That is the actual commercial cost, and it is why the instinct to fix it is correct even when the proposed fix is wrong.
Why a redesign usually makes it worse first
Here is the mechanism that catches people. A rebrand produces a new logo, a new palette and a set of new files. It does not produce the discipline that was missing. So the old assets do not disappear; they carry on living in decks, in email signatures, in the sales team's shared folder, in the PDF a customer downloaded in March.
What you now have is the previous fragmentation plus a second, incompatible system. For twelve to eighteen months, until every artefact has cycled through, the company looks less coherent than it did before the money was spent. Anyone who has worked through a rebrand recognises this period, and it is the reason a lot of second rebrands happen four years after the first.
The redesign also resets whatever recognition you had accumulated. Recognition is slow, expensive and boring to build: it comes from repetition, from someone seeing the same shape in the same colour enough times that it stops needing to be read. Throwing it away is sometimes correct, but it should be a deliberate decision with a stated reason, not a side effect of wanting the deck to look better.
The test
There is a straightforward way to tell which problem you have. Take your logo, your two main colours and your typeface. Apply them, correctly and consistently, to five things: your homepage header, one deck, one email, one social post and your invoice. Do it properly, once, by hand.
Then look at the five together. If they now feel like one company and the only complaint you can make is that the aesthetic is not to your taste, you have a consistency problem. The assets were fine; the application was not. If, applied correctly and consistently, the result still communicates the wrong thing, says nothing about who you are for, or actively misrepresents what you sell, then you have a design problem and a redesign is warranted.
Most teams who run this test discover the first answer. That is not a comfortable finding, because the fix is less exciting than a rebrand and harder to celebrate internally. Nobody announces a consistency programme on LinkedIn.
What the cheaper fix involves
Fixing consistency is not a design project in the usual sense. It is closer to an inventory exercise followed by a small amount of manufacturing.
First, decide the things that were never decided: one logo file and one set of lockups, two type sizes for headings and one for body, a primary and a secondary colour with defined jobs, and a spacing step. That is perhaps a day of real decision-making and it can be written on two pages.
Second, build the templates that make following those decisions the fastest route. This is the part that does the work. A deck template with real master slides, a social template set with the type sizes already in place, an email template, a document template. A rule that exists only in a PDF gets ignored under deadline; a rule that is already loaded into the file people open gets followed without anyone thinking about it.
Third, remove the old assets from circulation. Delete them from the shared drive rather than moving them to an archive folder, because an archive folder is just a slower shared drive. Replace the email signatures centrally. Update the templates the sales team actually uses rather than the ones they were told to use.
Done properly, this costs a fraction of a rebrand and produces a visible change within weeks rather than months, because you are not waiting for new artwork to propagate. You are making the existing artwork behave.
When a redesign really is the answer
To be clear, there are cases where new artwork is the correct spend. If the mark is genuinely unusable at the sizes your product requires, that is a drawing problem and no amount of discipline fixes it. If the identity was made for a business you no longer run, if you have moved from selling to consumers to selling to enterprises, the old design is accurately representing a company that no longer exists. If there is a trademark conflict, the decision has been made for you.
The useful discipline is to state which of those applies before committing. Write one sentence: we are redesigning because ___. If the sentence you end up with is some version of it feels dated, go back and run the five-artefact test first. The odds are good that you will save most of the budget and get a better-looking company faster than the rebrand would have delivered one.