How often should you refresh your brand strategy?

Every eighteen months or so, a familiar conversation starts. The brand feels tired. The deck looks like it was made in a different era, which it was. Someone says the words "we've outgrown this," everyone nods, and a rebrand gets scoped.

Six months and a considerable budget later there is a new logo, a refreshed palette, a tagline, and a set of slides. Sales are roughly where they were. Nobody says this out loud, because the new one does look better.

What happened is that a real problem got treated with the wrong instrument. The brand probably had genuinely outgrown something. Almost certainly not its logo.

This piece uses the three-layer model - vision, expression, perception. If you have not read [what AI can and can't do in brand strategy](/notes/what-ai-can-and-cant-do-in-brand-strategy), that is where the layers are explained, and it is the shorter way in.

The short answer

Vision changes every three to five years, or when the business genuinely changes. Positioning is reviewed quarterly and changed only when the evidence diverges two reads running. Expression changes continuously. Putting all three on one annual cycle is what produces expensive rebrands that fix nothing.

The rest of this is how to tell which layer is actually asking for attention.

Brand refresh vs rebrand: the difference that matters

These get used interchangeably and they are not the same decision.

A rebrand changes what the brand claims to be. New position, sometimes a new name, usually a new customer. It is a strategic move and it resets whatever recognition you had built, which is why it is rare and why it should be evidence-led.

A brand refresh changes how an unchanged claim is expressed. Same position, better carried. New site, sharper copy, a visual system that has caught up with the product.

The reason the distinction matters is cost. A refresh is regular maintenance. A rebrand spends accumulated recognition and asks the market to learn you again, and most brands that think they need one need a refresh with a clear head.

The test is one question: has what we actually mean changed, or only how well we are saying it? If it is the second, you are not rebranding. You are doing your job.

Vision: every three to five years, or a real business change

The vision layer is what the founder believes the brand is, means and promises. It should be the stillest thing you own.

Change it when the business has actually changed:

• A new product category that the current statement cannot stretch to cover
• A new market with different conditions and a different buyer
• A founder change, or a change in who makes the decisions
• A sustained shift in who is buying, where the customer you now have is not the customer you built for

Notice that every item is an event, not a date. There is no calendar reason to revisit a vision. If the business has not changed and the vision no longer feels right, that feeling is worth examining, but it is more often boredom than strategy. You are exposed to your own brand hundreds of times more often than your customer is. Being tired of it is not evidence.

The rare exception is when market data suggests the founder's framing does not match a real customer need. That is a genuine vision problem and it is serious. Do not start expression work until you have tested it, because expressing a wrong idea better just makes it wrong faster.

Positioning: review quarterly, change on evidence

This is the layer people get wrong in both directions, either never revisiting it or rewriting it whenever someone joins.

Reviewing is not changing. A quarterly review means putting three things side by side: your positioning statement, your best-performing creative, and the language customers actually used about you in the last ninety days. Then asking whether they are still telling the same story.

Most quarters, they are. You note it and move on, and the review takes an hour.

Change when the same divergence shows up twice running. One quarter of odd data is noise: a seasonal campaign, a viral post, one loud customer segment. Two consecutive quarters showing the same drift is a pattern, and patterns earn a response.

This discipline is the entire defence against the two failure modes. Reviewing often stops you drifting for two years without noticing. Requiring evidence twice stops you rewriting your positioning because of one bad month.

Expression: continuously

The expression layer should never be still. Copy, creative, channel, format, offer - all of it in motion, all of the time.

That is not chaos, it is a loop with rules:

• Three to five live hypotheses at once, no more
• Each one written as a specific prediction with a metric and a window attached
• No more than three variables changed across everything running
• Seven to fourteen days minimum before reading a result
• Every outcome recorded, including the failures, because an unrecorded test is an expense rather than an investment

Run that properly and something interesting happens to the question this article is about. Expression work generates the evidence that the quarterly positioning review needs. The layers are not independent - the fast one is what tells you whether the slow ones are still right.

Brands that do not run this loop have no evidence at review time, so they end up deciding on instinct, which is how you get an annual rebrand in the first place.

Five signs you need a brand refresh now

Forget the calendar. These are the signals that actually mean something is wrong.

1. Customers describe you in words you never chose. Read your last fifty reviews and list the recurring adjectives. If that list and your positioning statement are describing different companies, the market has settled on a version of you that you did not write.

2. Your best-performing creative contradicts your positioning. This one is uncomfortable and it is the most useful on the list. If the ad that works says something your strategy says you are not, one of the two is wrong, and the market has already voted.

3. A competitor has claimed your sentence. Positioning is relative. If someone else now owns the thing you were going to be known for, and owns it more loudly, your position has changed whether or not your document has.

4. The people buying are not the people you built for. Not automatically a problem - sometimes it is a better business than the one you planned. It is only a problem if your marketing is still addressed to the original group, in which case you are paying to talk past your actual customers.

5. There is a product line your current statement cannot accommodate. When you find yourself writing copy that quietly ignores the brand strategy because the strategy does not fit what you are selling now, the strategy has expired.

One of these means look closer. Two or more means the work is already overdue.

If two or more of those landed, the [Founder Decode pack](/notes/founder-decode) is the next step. It is free, it takes an afternoon, and it ends with the three columns side by side so you can see exactly which layer has drifted.

What a brand refresh is not

It is not a logo. A logo is an identifier, and identifiers get value from consistency rather than from quality. Changing it spends recognition and returns very little.

It is not a palette.

It is not a tagline that nobody outside the company will ever say aloud.

A refresh means the expression layer has caught up with what the brand already means, at every point where a customer meets it. That is mostly language, mostly unglamorous, and it does not photograph well for a case study. It is also the only version of this that moves anything.

So: how often?

Vision, hardly ever. Positioning, reviewed four times a year and changed when the evidence insists. Expression, always.

The honest answer to "how often should we refresh" is that a brand run properly never needs a refresh, because nothing has been allowed to drift far enough to require one. The refresh is what you do to recover from not watching. The watching is cheaper.

When was the last time you put your positioning statement, your best ad and your last fifty reviews on the same page?

That exercise is the whole of the Founder Decode pack, which is free and which you can run yourself this week.