The Brand Had Customers

Sixty-six percent of Americans viewed Sears favorably in August 2017. The retail industry average was 52 percent.¹

Sears had been a household name for 131 years. It had been the largest retailer in the country within living memory. Revenue had declined from $53 billion to $16.7 billion over the previous decade, a 68 percent drop.² The favorability score did not register the decline.

Fourteen months after the survey, Sears filed for bankruptcy.³

Toys "R" Us sold one in five toys purchased in the United States. Brand recognition among American parents approached 99 percent.⁴ Same-store sales had declined 3.4 percent over the holidays and 7 percent domestically in the final months before filing. The company entered bankruptcy in September 2017 with $5 billion in debt.⁵

BlackBerry defined the smartphone. In 2009, it held more than half the US market and 85 million subscribers worldwide. Revenue peaked at $19.8 billion.⁶ By 2016, market share had fallen below one-tenth of one percent. Revenue declined 97 percent, to $560 million.⁷

Bed Bath & Beyond operated 955 stores across the country. In 2022, 89 percent of American home shoppers recognized the brand.⁸ In April 2023, the company filed for bankruptcy. By July, every store was closed.⁹

Four companies. Each had above-average or near-universal brand recognition in the years before collapse. In every case, the awareness score stayed high while the structural position underneath it disintegrated. Revenue declined, market share evaporated, and stores closed. The metric that said the brand was strong was the last number to move. Solo Stove's Snoop Dogg campaign produced 52 million views and an Ad Age Top 40 ranking; the CEO was fired before the year ended.¹⁰

Brand awareness measures whether someone has heard of you. It does not measure whether they will buy from you, stay with you, or choose you over the alternative the distribution owner just placed beside you at 20 percent less.

The metric was the last number to move.

The global market research industry generates $142 billion a year.¹¹ Brand tracking accounts for roughly 7.5 percent of it. Corporations run continuous tracking studies modeled on the methodology Millward Brown built in 1976 for Cadbury Schweppes.¹² The question has not changed in 50 years: have people heard of us?

The answer, for every company that collapsed, was yes.

What determines the outcome

If brand awareness predicted survival, Sears would still be open. If it predicted market share, BlackBerry would still make phones. If it predicted revenue, Toys "R" Us would still sell one in five toys.

A 2011 meta-analysis in the Journal of Marketing Research examined 751 short-term and 402 long-term advertising elasticities across 56 studies spanning five decades of data.¹³ The mean short-term effect of advertising on sales was 0.12. The long-term effect was 0.24. Nearly half the elasticities, 47 percent, were not statistically significant.¹⁴

The advertising was real, the impressions were counted, the awareness was generated. The sales effect, in 47 percent of cases, was not measurable.

Price elasticity across the same consumer categories averaged −1.76.¹⁵

Price moved purchasing decisions 15 to 59 times more than advertising. The shelf price changed what people bought. The campaign changed whether people had heard of the brand. These are different measurements. The industry that tracks one spent $142 billion last year. The variable that determines the other is controlled by whoever owns the shelf.

Seventy-six percent of purchase decisions in grocery are made inside the store.¹⁶ Not before the drive. Not during the commute. Inside, at the shelf, in the aisle. Eye-level placement generates 23 to 35 percent more sales than the shelf below.¹⁷ The shelf selects which products the customer sees. The customer selects from what survived the filter.

If awareness determined the outcome, private label would not exist. Store brands carry no national advertising budget. Most consumers cannot name the manufacturer. No spokesperson, no Super Bowl commercial, no tagline, no brand awareness score. The product arrives on the shelf and sells because the shelf is there and the price is right.

Private label now holds 21.3 percent of US grocery by dollar share and $282.8 billion in annual sales. Both numbers are records. The share rises every quarter.¹⁸ Dollar sales grew 3.3 percent in the most recent period while national brands grew 1.2 percent. Unit share hit 23.5 percent, also a record. Private label is gaining in dollars and in volume simultaneously. The national brand is losing both.

In 2022 and 2023, inflation created a natural experiment. Consumers faced the same products on the same shelf at higher prices. National brand awareness did not decline during the period. If anything, the brands were more visible. Advertising continued. Campaigns ran. Brand tracking studies returned the same scores they always returned. The brand was recognized. The brand was favorably perceived. The brand was trusted.

Consumers switched to store brands anyway, not because they forgot the brand but because the shelf offered an alternative at a lower price and the quality gap had closed to the point where the choice was obvious. The awareness the brand had built over decades of advertising did not hold the customer when the price difference became visible on the same shelf.

Simon-Kucher, the global pricing consultancy, studied the shift and called it permanent: "The surge of private label purchasing during last year's inflationary recovery period wasn't a fluke. It was a reset."¹⁹

They did not switch back. Forty-nine percent of households earning more than $100,000 a year now predominantly buy private label.²⁰ The Freakonomics podcast documented that doctors and chefs were more likely to buy store brands in their own professional fields and feel proud about it, treating the choice as an expression of expertise rather than a concession on quality.²¹

The retailer's incentive to promote its own brand is structural. National brand margins average 26 percent. Private label margins average 35 percent.²² Every percentage point of shelf share that shifts from national brand to private label increases the retailer's profit. The brand rents space in a system that profits from replacing it.

The brand's awareness score did not decline during the inflation period. The price changed. The shelf offered an alternative. The loyalty evaporated in 18 months under a single variable: price. The awareness score, the one the brand tracked, did not register the shift. The distribution data, the one the retailer tracked, registered it immediately.

Two numbers. Both available. Both accurate. One measured recognition. The other measured the relationship. Nobody put them next to each other.

The landlord

Kirkland Signature generated approximately $90 billion in revenue in fiscal 2025.²³ The Coca-Cola Company, the most advertised brand in the history of consumer products, generated $47 billion.²⁴

Kirkland runs no national advertising. It has no mascot, no jingle, no celebrity endorsements, no social media strategy. It exists because Costco controls the shelf and decided to put its own product on it.

Costco has 145 million cardholders.²⁵ Its renewal rate is 92.9 percent. It caps markup at 14 percent on national brands, 15 percent on Kirkland. Kirkland products are priced at least 20 percent below the national brand equivalent. The policy is explicit.²⁶

When the national brand and the Kirkland version sit side by side on the same shelf, the customer follows the price. The awareness the national brand spent decades building does not overcome a 20 percent price gap in a store where the customer trusts the retailer more than the brand. The 92.9 percent renewal rate belongs to Costco. Not to the brands inside it.

Costco can replace any national brand with a Kirkland version at any time. It has done so across batteries, olive oil, vodka, diapers, coffee, laundry detergent, and prescription medication. The process is the same each time. Costco identifies a category where the national brand charges a premium. It sources a comparable product, often from the same manufacturer that makes the national brand. It places the Kirkland version beside the national brand at 20 to 30 percent less. The customer, who entered the store loyal to Costco and not to the brand, follows the shelf.

The national brand loses because the shelf was rearranged, not because it was forgotten. The awareness score does not register the change. The distribution agreement does.

The brand rented access to someone else's customer. The rent was revocable.

Amazon runs the same mechanism with better surveillance.

In April 2020, the Wall Street Journal reported that Amazon employees used third-party seller data to develop competing products.²⁷ What sells. At what price. Which search terms drive discovery. What the return rate reveals about product weaknesses. The practice was described internally as standard operating procedure. Employees used workarounds called "going over the fence" to access individual seller reports.

Jeff Bezos testified before the House Judiciary Committee three months later. Asked whether Amazon's stated policies prohibited using individual seller data to launch competing products, he said: "I can't guarantee you that policy has never been violated."²⁸

At its peak, Amazon operated more than 400 private-label brands across 243,000 products.²⁹ Amazon Basics batteries captured 31 to 33 percent of online battery sales. Duracell, the brand with the copper top and four decades of television advertising, was second at 21 percent.³⁰

The brand funded its own replacement's market research.

The platform did not just control the shelf. It watched what sold, learned why, and built the replacement using the brand's own sales data. The brand's investment in reaching customers through the platform was the platform's product development pipeline.³¹ The brand paid Amazon for advertising to reach customers on Amazon. The advertising data told Amazon which products customers searched for, which features they valued, which price points triggered conversion. Amazon used that data to build the competing product. The brand paid for the intelligence that was used against it.

The creator's followers belong to the platform. The brand's customers belong to the retailer. The mechanism is the same. The distribution owner controls the relationship. The person who believes they built the audience discovers the audience was never theirs.

Spotify determines which artists reach listeners through playlist placement. An artist placed on a major playlist can see streams increase by orders of magnitude overnight. An artist removed sees them collapse just as fast.³² The artist believes they have fans. The playlist has listeners. When the playlist changes, the audience changes with it. The artist's "brand" did not decline. The distribution changed. The fans followed the distribution.

In 2012, a brand's Facebook post reached roughly 16 percent of its followers. By 2025, organic reach had collapsed to between 1.4 and 2.6 percent.³³ An 84 to 91 percent decline over 13 years. The brand spent years building 500,000 followers. It ran contests, posted content, responded to comments, invested in a community. The followers were Facebook's users, shown the brand's content at Facebook's discretion. When the algorithm shifted to prioritize Reels and paid placement, the audience the brand believed it had built disappeared. The follower count did not change. The reach did. The brand still had 500,000 followers. It could reach 7,000 of them.

Four distribution owners. Costco controls the shelf. Amazon controls the search result. Spotify controls the playlist. Meta controls the feed. In each case, the brand rented access to someone else's customer and measured awareness while the structural relationship belonged to whoever controlled the last interaction before the purchase.

170 years

The belief that the brand owns the customer was formed in one specific era, lasted roughly 40 years, and has been wrong for the 60 years since.

The distribution owner changed five times. The belief did not.

1850s to 1910s. The shopkeeper decided what to stock. The customer trusted the person behind the counter. The brand was whatever the shopkeeper recommended. A general store in Ohio in 1880 might carry three kinds of flour. The customer asked which one was good. The shopkeeper answered. The manufacturer in Minneapolis had no way to reach that customer, no way to influence the recommendation, no way to build a relationship that bypassed the person behind the counter. The shopkeeper owned the customer.

1920s to 1960s. Radio, then television. For the first time in the history of commerce, the brand could reach the consumer without the retailer's permission. Procter & Gamble built the mechanism. The term "soap opera" is literal. P&G's first serial drama, Ma Perkins, launched on radio in 1933.³⁴ By 1935, P&G was the largest buyer of network radio time in the country, sponsoring 778 program hours in a single year. By 1937, the company spent $4.5 million on advertising, 90 percent of it on daytime programming.³⁵

The brand could talk to the customer directly. For the first and only time, the brand owned the relationship.

By the 1950s, a prime-time television commercial could reach 70 percent of American households in a single evening. Three networks controlled the signal. The cost of a national campaign was enormous, which was the point. Only large companies could afford it. The expense itself was the barrier, and the barrier created concentration. The brands that could pay for national television became the brands that every household recognized. Procter & Gamble, General Mills, Colgate-Palmolive, Kraft. The retailer could not compete with that reach. For 40 years, the manufacturer had direct access to the living room, and the living room had direct access to the brand. The brand believed it owned the customer because it could reach the customer without anyone's permission. The belief was accurate. For that era.

This era created the measurement. Daniel Starch built the first advertising recognition test in 1923. George Gallup developed aided recall the same year.³⁶ Millward Brown launched the first continuous brand tracking study in 1976.³⁷ Every innovation in measuring brand strength was built to capture what mass media made possible: recognition at scale. The belief and the metric were born in the same era. They aged together. Neither was updated when the conditions changed.

1970s to 2000s. Big-box consolidation. Walmart. Target. Costco. Distribution concentrated into fewer, larger retailers with more purchasing power than any manufacturer. Slotting fees emerged in the early 1980s. By 2000, 85 percent of retailers charged them.³⁸ The estimated cost to introduce a four-item product line across all US supermarkets reached $16.8 million.³⁹ The brand now paid for the right to be on the shelf.

The relationship had shifted back to the distribution owner. Walmart decided which brands reached 140 million weekly shoppers. Target decided which end caps got traffic. Costco decided which brands entered a warehouse where the average member spent $170 per visit. The brand paid the fee. The brand ran the campaign. The brand measured the awareness. The retailer controlled the shelf. The brand's memory of the mass media era, the 40 years when it could reach the customer directly, made the rental feel like ownership. The metric confirmed the feeling. People had heard of the brand. The metric was measuring recognition. The brand was assuming relationship.

2010 to 2018. DTC. Shopify, Instagram, Facebook ads. Direct access to the consumer at scale. The cross-category average customer acquisition cost was $9.⁴⁰ The DTC brands believed they had cut out the middleman and owned the relationship directly. By 2022, the average had risen to $29. A 222 percent increase, averaged across every category, and that was if you could get the customer at all.⁴¹ In fashion and beauty, the real number was $100 to $200 per customer. Some brands spent more acquiring the customer than the customer's first order was worth. The relationship had always been rented. The rent was paid to Facebook, Google, and Instagram instead of to Walmart and Target. The landlord changed. The rental did not.

2018 to present. The platforms consolidated. Amazon, Google, TikTok, Instagram. The brand rents access again. The new distribution owner has better data than any previous landlord. It controls search, recommendation, ad placement, and behavioral data simultaneously. The brand rents access to a customer the platform understands better than the brand does.⁴²

The belief was formed in an era when it was briefly true. The era ended. The belief didn't.

170 years. Five eras. The distribution owner changed every time. The brand's belief that it owned the customer was formed in era two and never updated after the conditions changed.⁴³ The metric that confirmed the belief, brand awareness, was invented in the same era. It tracked recognition, not the relationship, not the structural position. The brand carried the belief from the era of the television commercial into the era of the algorithm. The television is off and the algorithm is on, but the belief is unchanged.

What ownership looks like

The brands that survived the landlord own their own distribution.

Hermès generates approximately 95 percent of its revenue through its own retail network. 310 boutiques worldwide. No wholesale dependency. No intermediary determines which customer sees which product.⁴⁴ The distribution is the brand.

Hermès does not need to measure whether people have heard of it. The waitlist is the measurement. Estimated production of 12,000 to 15,000 Birkin bags per year. Wait times of one to three years.⁴⁵ The scarcity is structural, not manufactured. The relationship belongs to Hermès because no intermediary routes the customer to a competing product at 20 percent less.

Apple operates more than 530 stores across 27 countries. Thirty-seven percent of its revenue flows through direct channels, up from 29 percent in 2018. The share is growing every year.⁴⁶ Apple's brand awareness is among the highest in the world, but it is not the survival mechanism. Ninety percent of iPhone owners buy another iPhone. Eighty percent own multiple Apple devices.⁴⁷ The switching cost is the ecosystem: photos stored in iCloud, messages locked in iMessage, apps purchased through the App Store, health data synced across Watch and iPhone. Leave Apple and you leave the infrastructure of your digital life. The carrier that once controlled the phone upgrade decision has been bypassed. The awareness is a consequence of the position, not the cause.

Tesla sells directly in 30 states. No dealers. 276 locations. After the sale, the relationship continues. Over-the-air software updates change the car without a service visit. 75,000 Supercharger connectors across 54 countries create a fueling network the brand controls. 8.4 billion cumulative miles of driving data feed the development of the next product.⁴⁸ Every Tesla on the road is collecting information for the company that sold it. The brand that owns the distribution owns the data. The brand that owns the data owns the relationship. The dealership model, which every other automaker uses, places a third party between the manufacturer and the driver. The dealer owns the customer data. The dealer controls the service relationship. The manufacturer built the car and handed the customer to someone else.

In every case where the brand survived, the mechanism was the same. The brand owned the distribution, the customer data, and the relationship.⁴⁹ Awareness was a consequence of the structural position, not the cause of it. Hermès survives because no intermediary stands between the boutique and the customer. Apple survives because leaving the ecosystem costs more than staying. Tesla survives because the data flows to Tesla, not to a dealer.

The brands that rent distribution rent the customer. The awareness score says the brand is known. The distribution agreement says the relationship belongs to someone else.

$142 billion is spent each year on market research. Brand tracking is 7.5 percent of it. The methodology has not changed in 50 years. The question has not changed: have people heard of us?

Sears. Yes. Toys "R" Us. Yes. BlackBerry. Yes. Bed Bath & Beyond. Yes.

The metric confirmed what the company wanted to hear every quarter it was run. The structural position, the one that determined whether the company would survive, was in the distribution agreements, the shelf allocation data, the algorithm's weighting, the platform's private-label pipeline. Nobody put the awareness number and the distribution number next to each other. Nobody compared the map to the territory.

Sears had 66 percent favorability while revenue fell 68 percent. Toys "R" Us had 99 percent recognition while same-store sales fell 7 percent. BlackBerry had 85 million subscribers while the smartphone market moved to a platform it did not own. Bed Bath & Beyond had 89 percent awareness while every store closed. The brand awareness survey, run quarterly, confirmed the belief quarterly. The quarterly earnings report, also run quarterly, contradicted it. Both documents were on the same executive's desk.

170 years. Five eras. The distribution owner changed every time. The brand's belief that it owned the customer was formed when mass media briefly made it true, installed into corporate culture while television was still the primary medium, and never audited after the conditions changed. The metric that confirmed the belief was invented in the same era and tracked the same thing. Recognition, not the relationship, not the structural position, not who owned the last interaction before the purchase.

The brand narrative was the philosophy. Distribution was the budget. The budget determined the outcome.⁵⁰

The brand had customers. The customers were never the brand's.