Rebranding: when to change your brand and how to measure it

A rebrand is justified when the brand no longer describes what the company sells, who it sells to or where it sells. Branding is the work of building and managing the brand in buyers' memory; rebranding means redoing part of that work. The legitimate reasons are strategic and verifiable; leadership fatigue, a new CEO or the urge to look modern are not among them.

The signals that justify opening the project:

• Merger or acquisition. Two brands became one company and the brand architecture, meaning the relationship between the masterbrand, product brands and sub-brands, became confusing to customers.
• Change of business. The company moved into categories its current name or positioning doesn't cover, and the brand started holding back sales of the new offering.
• New market. The name is hard to pronounce, is already registered by another company or carries a bad connotation in the target country.
• Damaged reputation. The brand carries a crisis that operations have already fixed, and research shows the negative association persists.
• Lost differentiation. In research, buyers can't tell the brand apart from competitors, and the company competes on price alone.

The false signals are internal. Your team sees its own brand every day and tires of it long before buyers do, who see it in passing, among dozens of others. A competitor changing its identity isn't a reason either. If none of the five signals above shows up, the answer is usually a refresh: adjusting design and application without replacing the elements the market recognizes.

Distinctive brand assets are the elements that make buyers think of the brand without reading the name: color, symbol, packaging, character, sonic logo. They support mental availability, the likelihood that the brand comes to mind at the moment of purchase. These shortcuts take years to build, and they are rare: only 15% of the assets Ipsos tested are truly distinctive².

Be Distinctive. Everywhere., the study by Ipsos and the agency JKR, tested logos, slogans, mascots, colors and products with over 26,000 respondents worldwide². The global results by asset type show the logo is not the strongest asset: product shape reaches 31% of assets at the top standard, logos 19%, slogans 6% and color on its own 4%².

To decide what to keep, the Ehrenberg-Bass Institute, the marketing research center at the University of South Australia, uses the Distinctive Assets Grid developed by Jenni Romaniuk. It crosses two axes: fame (how many category buyers link the asset to the brand) and uniqueness (how many link it to that brand alone). The reference cutoff is 50% on both axes³. An asset that clears both cutoffs should survive the rebrand, even if redrawn.

The public cases with disclosed numbers share a pattern: the company replaced an element buyers used to recognize the brand, without first measuring what it was worth. The figures below come from the US and Europe and were reported by the press based on company data and research firms. Brazil, where this desk is based, has no comparable public case with before-and-after data.

Tropicana is the textbook case. The Pure Premium line was bringing in more than 700 million dollars a year, and the launch campaign for the new packaging cost 35 million dollars⁷. The new carton removed the orange with a straw, the asset shoppers used to find the product on the shelf. The juice was the same; the visual identity changed, and shoppers stopped recognizing the brand in the aisle.

At Gap, the backlash came before any sales data. Beyond the comments, a protest Twitter account gathered 5,000 followers and a parody site collected almost 14,000 versions of the logo⁴. The company reversed course without measuring the sales effect, which is also a mistake: it decided based on noise, with no baseline to compare against.

Cracker Barrel had data. A YouGov poll of 1,000 US adults, taken the weekend after launch, found 65% awareness of the new logo and 76% preference for the old one⁸. When the chain announced it would bring back the old logo, the stock rose more than 8%⁸. At the low point of the launch-day session, the loss in market value reached almost $200 million⁵.

The Jaguar case shows the risk of blaming everything on the new identity. In April 2025, the brand registered 49 cars in Europe, against 1,961 a year earlier, and was down 75.1% from January to April⁶. At the same time, the company had discontinued almost every model while waiting for its EVs. The drop mixes a lack of product with a reaction to the brand.

The timeline is documented. Jaguar stopped selling new cars in the UK in November 2024 and ended production of the F-Pace, which the year before had been the brand's global best-seller, with 21,943 units⁹. With almost nothing to sell, dealers couldn't register cars either way, whether or not the public liked the new campaign.

Without a brand baseline measured before the change, no one can separate the two causes. To do so, Jaguar would need metrics that don't depend on inventory: branded search, purchase consideration, preference among luxury car buyers. That is the core argument here: a rebrand without prior measurement can't prove it worked, or that it failed.

Measure across three horizons. In the first weeks, sentiment and recognition show whether buyers can still find the brand. Over quarters, sales, market share and branded search show whether demand held up. Over years, preference and price premium show whether the brand got stronger. Long-term measurement is what justifies the investment.

The reason to take the long term seriously lies in brand value research. In Kantar BrandZ, the portfolio of brands with the highest brand equity returned +88% relative to the S&P 500 and +251% relative to the MSCI World Index¹⁰. That is a stock market correlation, not proof of cause, but it shows what is at stake when a rebrand weakens the brand.

Always separate short-term from long-term effects. Negative reaction in the first weeks may fade as people get used to the change; a sales decline that persists for quarters rarely does. In Tropicana's 2024 case, sales fell 8.3% in July, 10.9% in August and 19% in October¹, and the new bottle was also smaller, which mixes rejection of the design with a reaction to price.

Before approving any design, record where the brand stands today, using the same instruments you will use afterward. Without that snapshot, the rebrand turns into a debate about taste, and the company decides based on social media noise, as Gap did. The checklist below is what we recommend for a mid-size or large company.

1. 01
   Write down the reason. In one sentence, state which strategic signal justifies the change and which metric it should improve. If you can't write the sentence, the project doesn't have a reason yet.
2. 02
   Inventory your assets. List color, symbol, packaging, character, sonic logo and tagline. Test each one without the brand name among category buyers and flag the ones that clear the fame and uniqueness cutoffs.
3. 03
   Freeze the baseline. Record share of search, sales, market share, preference, consideration and average price over the past twelve months.
4. 04
   Decide what stays. Strong assets get redrawn, not discarded. Weak ones can go.
5. 05
   Test before launch. Show the new design to real buyers in a buying context, on the shelf or on screen, and measure whether they can still find the brand.
6. 06
   Roll out in stages if you can. One region, channel or product line first, with a control group for comparison.
7. 07
   Schedule the reviews. Agree with leadership on dates to read the metrics at weeks, quarters and years, and on the criterion that would trigger a reversal.