The client always arrives the same way. Sales are soft and somebody has already suggested a new logo. They want a makeover. What they need is a detective, because something in the business has died quietly and nobody can say when, or who was holding the knife.

A brand audit is a structured investigation that compares what a company says about its brand, what it actually delivers and what customers believe, then traces every gap back to its cause.

So work it like a case. Collect the physical evidence. Question the witnesses. Test the alibis.

Then gather everyone in the drawing room and name the culprit. It is rarely the one the client suspected.

Every case starts with a body

Nobody commissions an audit when things are going well. The phone rings when revenue slips, when a competitor starts eating your lunch, when a merger leaves two brands sharing one bed, or when the founder notices that the people in the shop no longer look like the people in the brochure.

Krispy Kreme is a file worth keeping on the desk. The company floated in April 2000, and by early 2004 it had almost 400 stores. It also sold doughnuts through supermarkets, petrol stations and convenience stores, a channel that at one point brought in up to half its sales.

A treat that once justified a special trip became something you grabbed next to the motor oil. The novelty went.

In May 2004 the company missed its quarterly estimates for the first time as a listed business, and the chief executive blamed the low-carb diet craze. By 2005 the shares had lost three-quarters or more of their value and regulators were going through the accounts.

Low-carb diets were the convenient suspect. Ubiquity was the likelier one. First rule of any audit: the explanation the client brings you is a statement, not a fact.

Some cases get reopened. In March 2024 Krispy Kreme announced plans to put its doughnuts in every McDonald’s in the United States. By June 2025 the two were parting ways, citing unsustainable operating costs and thin demand. Same scene, different decade.

Secure the scene: the brand inventory

Before you talk to anyone, walk the scene and bag everything. Every place the brand shows up: signage, packaging, website, social accounts, invoices, uniforms, hold music, the reseller’s van, the email signature somebody in accounts designed in 2017, and the product range itself.

Lay it all out on one table. Photograph it. You are looking for what investigators call a pattern and what clients call “a bit inconsistent”.

When Angela Ahrendts became Burberry’s chief executive in July 2006, that table would have been crowded. The brand had 23 licensees around the world, each doing something different. The famous check, once mostly a coat lining, had been splashed across everything, and cheap counterfeits spread it further. Between 2001 and 2005 it had become the unofficial uniform of British “chav” and football-casual culture, to the point where some venues turned away anyone wearing it.

You do not need a lab to read that scene. A luxury brand that appears everywhere stops being luxury. Burberry pulled the check back to roughly a tenth of its products, bought out its Spanish franchise, brought its fragrance and beauty licences in-house and rebuilt around the trench coat. During Ahrendts’ tenure, which ended when she left for Apple in 2014, sales passed £2 billion.

Burberry rebuilt itself around the trench coat, the product its own executives had stopped wearing.

Muji’s scene told a quieter story. Born in 1980 as a no-brand private label for the Seiyu supermarket chain, it had grown by 2001 to more than 250 outlets and over 5,000 items, some of them cheap but substandard. Kenya Hara, its new art director that year, judged that the company had stalled around its founding idea.

Thousands of products, one principle, and the principle was disappearing under the stock. An inventory makes that kind of drift visible in a single afternoon.

Question the witnesses

Evidence tells you what the brand puts out. Witnesses tell you what lands. Interview them separately, and never in the client’s boardroom, where everyone remembers the official version.

Start with customers, though not the loyal regulars who answer every survey. Find recent buyers, lapsed buyers and the people who considered you and walked away. Then talk to the front line, who hear the complaints management never sees, and to distributors, suppliers and agencies, who see your brand from the side. Question the leadership last, and with suspicion.

Sometimes the most damning witness is on the payroll. At Ahrendts’ first strategy meeting at Burberry, held in London on a grey, damp day, not one of her top 60 executives was wearing a Burberry trench coat. No survey required. When your own people will not wear the product, the statement writes itself.

Witnesses can also hand you something you never expected. Indomie, the Indonesian instant noodle, reached Nigeria in 1988 and has been made there since 1995, when Dufil Prima Foods, a joint venture between Indofood and Singapore’s Tolaram Group, opened the country’s first instant noodle plant.

It is now so woven into daily life that many Nigerians believe it is a Nigerian brand. An audit that turned up that belief would not log it as a problem. It would log it as an asset under protection.

Indomie’s Nigerian advertising speaks so local that many assume the brand was born there.

How do you find out what customers really think?

Not by showing them your logo. Strip a logo of its context and people can only tell you whether they like the colour, which is a taste test, not an interrogation.

Ask about behaviour and memory instead: when they last bought, where, instead of what, and why.

Write down their exact words. The language customers use is evidence in its own right.

Then read the unprompted material: reviews, complaints, comment threads, search queries, the questions people put to your sales team. None of it was written to please you, which makes it more reliable than anything collected on your letterhead.

Mystery shopping helps, with a caveat. Staff who suspect a check put on a performance, so visit unannounced and more than once. The second visit is usually the honest one.

Be ready to hear the worst. In 2009 American consumers rated Domino’s the worst-tasting of the national pizza chains, level with Chuck E. Cheese. At the end of that year the company did what most brands never would.

It filmed customers criticising its pizza, ran the footage as advertising and changed the dough, the sauce and the cheese. The confession became the campaign. That only works if you have actually listened first.

Domino’s turned its harshest customer feedback into the campaign itself.

Check the alibis

Every brand carries an alibi: the story it tells about itself. In practice there are at least four versions of any brand. What it actually is, what it says it is, what people believe it is and what it has promised.

The audit lines them up side by side. Where they drift apart, you have a motive.

Abercrombie & Fitch had a famous alibi: exclusivity. Under Mike Jeffries, chief executive from 1992 to 2014, the stores were dark, loud and heavy with fragrance, with shirtless models at the door. In a 2006 interview he said the company went after the cool kids.

The alibi collapsed in public. In 2011 the company offered cast members of the reality show Jersey Shore money to stop wearing its clothes, and in 2013 the old interview resurfaced and the backlash arrived.

Jeffries resigned in December 2014. By 2015 the sexualised advertising was gone and the store “models” had become “brand representatives”. Fran Horowitz took over as chief executive in 2017.

Some alibis hold. Old Chang Kee says it has been making snacks since 1956, and the record backs it: the business began as a stall outside the Rex Cinema on MacKenzie Road and was bought by Han Keen Juan in 1986.

In January 2005 it won halal certification for its products, outlets and central kitchens, which told Muslim customers across Singapore that the curry puff was theirs too. A heritage claim with a checkable origin is an alibi you can put on the stand.

And some alibis are true but damning. Tata unveiled the Nano in January 2008 at ₹100,000, billed as the lowest-priced production car in the world. Accurate. Also fatal, probably.

A car is a statement about its owner, and “cheapest” is not a statement most buyers want parked outside the house. Tata had projected 250,000 sales a year; the best year managed about 74,500, and production stopped in 2018.

A sharp audit would have flagged the gap between what the brand said and what buyers wanted the car to say about them.

Follow the money

Detectives follow money because money does not lie about where it went. In an audit, that means sales by channel, product, region and customer type, set against where the brand spends its attention and budget.

Put the two columns next to each other. If the line that earns most of the margin gets a sliver of the marketing, or the newest venture eats half the management meetings while bringing in a fraction of revenue, you have found a motive.

Brands drift towards whatever is exciting, not whatever pays. Check the discounting too. A brand that is permanently on promotion is confessing something.

Lego’s books in the early 2000s were a crime scene. A burst of new ventures, from theme parks and video games to electronics, jewellery and education centres, had pulled the company away from the brick. By 2003 it was losing heavily and close to running out of cash.

Jørgen Vig Knudstorp, a former McKinsey consultant, became chief executive in 2004. The in-house video games division closed that year, the Duplo name returned and the Legoland parks were sold in 2005. Lego was back in profit the same year. The money had been pointing at the core all along.

Can you audit your own brand?

You can do the legwork: the inventory, the review-mining, the sales analysis. A small business should do it at least once a year.

But insiders make poor detectives on their own case. They see the brand they intended rather than the one they built, and nobody wants to tell the founder that the beloved tagline confuses everyone. Insiders also share the client’s suspect list, so they look hardest where they already expect trouble.

Picture a third-generation bakery in Toa Payoh that has always believed customers come for its heritage recipes. The owners run their own audit, ask the regulars and hear exactly what they expected.

An outsider interviews lapsed customers instead. The real draw, it turns out, was the 7am opening and the easy parking, both of which disappeared when the shop moved. The recipes matter. They are not the case.

Bring in an independent investigator when the stakes are high: before a rebrand, after a merger, ahead of a new market, or when the numbers have been sliding for a year and nobody can say why.

The reveal

At the end, gather the suspects. A good audit report is short enough to read in one sitting and blunt enough to act on. It sets out what the brand claims, what the evidence shows, where the two diverge and why, and what to do about it, in order of priority.

Show the exhibits: customer quotes in their own words, photographs of the rogue touchpoints, the sales chart that tells the story without help. Name the culprit plainly. A report that blames “a range of factors” has let everyone go home.

Expect resistance from whoever owns the quarterly numbers. Brand damage builds slowly and shows up late, so tie every finding to money someone already tracks: churn, discount depth, the price premium you used to command.

Then comes the hard part: keeping the client away from the design studio for another month. Burberry did not solve its problem with a new logo. Lego did not redraw the brick.

Both cut, refocused and made the brand true again, and design followed the verdict instead of replacing it. Diagnosis first. Surgery second.

Case notes: three quick answers

How often should a business run a brand audit?
A light review every year and a full audit every two to three years is a sensible rhythm. Run one immediately after a trigger event such as falling sales, a merger, a move into a new market or a planned rebrand.

What is the difference between a brand audit and a marketing audit?
A marketing audit examines activity: campaigns, budgets and channel performance. A brand audit examines meaning: what the brand stands for, how consistently it shows up and how customers actually perceive it.

What should a brand audit deliver?
A clear diagnosis of where claim, delivery and perception part company, the evidence behind each finding and a prioritised list of actions. Those actions may or may not include changes to the visual identity.

If there is a body on your floor and nobody can name the culprit, Firefish’s brand audit team takes cases.