We Got In

The customer chose from eight options. It looked like a market. It was a filter.

Cedric Atkinson

In 1806, a twenty-three-year-old from Boston loaded a hundred and thirty tons of ice onto a sailing ship and sent it to Martinique.

The ice came from a pond on his family's estate in Saugus, Massachusetts. The raw material cost nothing. Frederic Tudor's insight was not the ice. It was the infrastructure required to deliver it. He built insulated ice houses in port cities from Havana to Calcutta. He packed the blocks in sawdust, a waste product from New England lumber mills that also cost nothing. He negotiated freight on ships that were sailing south with empty holds. Over four decades, he extended the route to India: a sixteen-thousand-mile voyage that crossed the equator twice and took four months.1

The customer in Calcutta bought the ice. They did not know it came from a Massachusetts pond. They did not know that other ice cutters in New England had tried to ship and failed because they could not build the ice houses, insulate the holds, or absorb the losses on routes that took years to become profitable. Tudor was the only one who could afford to be there. The product was a commodity. The distribution was the filter.

By 1856, three hundred and sixty-three ships carried a hundred and forty-six thousand tons of ice out of Boston Harbor in a single year. Ice was the second-largest American export, behind only cotton.2 The customer never saw the filter. They saw the ice.

The product changed. The filter did not.

The customer walks into a store. Eight products on a shelf, lined up in a row. Different brands. Different prices. Different packaging. They pick one.

Eight options. The customer chose from what was available.

The shelf is not the market. The shelf is what survived the filter.

Thirty thousand new consumer products enter American stores every year.3 Within three years, more than half are gone.4 The customer never sees most of them arrive or leave. They see the shelf the day they walk in and assume it represents what exists.

What the customer does not see is what could not get there. The company with a better product and no capital for an entry fee. The startup that could supply a hundred stores but not two thousand. The manufacturer who could not survive ninety days without being paid. The products the customer might have preferred never reached the shelf.

The shelf has no label for what is missing.

The filter

To reach that shelf, a product had to answer five questions. The customer never hears the questions. They see the shelf after the answers have been graded.

$250,000

Can you pay to enter? Slotting fees, the price a manufacturer pays for shelf space, range from twenty-five thousand dollars per product in a regional grocery chain to a quarter of a million in high-demand markets.5 A national grocery launch for a single product can cost one and a half to two million dollars in slotting allowances alone.6 Costco does not charge slotting fees. Walmart did not charge them until 2015. Most grocery chains charge them on every new item that touches the shelf. The FTC has not updated its study on slotting allowances since 2003.7 The fees remain deliberately opaque. Nobody publishes the rate card. The system prices access to the customer, and the price is whatever the retailer decides it is.

Five per week

Can you sell fast enough? The system measures velocity: units sold per store per week. In refrigerated beverages, the threshold is five to seven units per store per week.8 Fall below it and the product disappears at the next category review. The customer does not attend the review. They see the shelf the week after.

Ninety days

Can you survive not getting paid? Major retailers pay on Net-60 or Net-90 terms. The product ships. The invoice goes out. Payment arrives two or three months later, minus deductions for chargebacks, markdowns, damages, and promotional costs that run five to fifteen percent of the total.9 The manufacturer finances production, packaging, and shipping for months before it sees a dollar.

In 2024, Jordan Buckner, the founder of a kombucha brand called FoodBevy, invoiced the national distributor KeHE for thirty-two thousand dollars. He received twenty-one thousand.

$32,000 Invoiced to distributor
$21,000 Received after deductions

Eleven thousand dollars, thirty-four percent of gross revenue, disappeared into chargebacks, fees, and returned products. The largest charge, eight thousand four hundred dollars, came from a single retailer returning unsold inventory. Buckner stopped selling through KeHE because continuing would have put his company out of business.10 The customer who bought the kombucha at Jewel-Osco had no idea the company that made it was being consumed by the cost of being there.

Two thousand doors

Can you supply two thousand stores at once? The product needs to be manufactured, packaged, labeled, and tracked through electronic data interchange, all at spec. At Target, an unscannable barcode incurs a seventy-five-cent penalty per carton, hundred-dollar minimum.11 Costco requires five million dollars in liability insurance before a product enters the warehouse.12 The precision is industrial because the system is industrial.

Fifty-three dollars

Does your product leave enough room for the retailer to make money? On general merchandise, major retailers take thirty-five to fifty percent of the retail price.13 On a hundred-dollar product, after the retailer's cut, payment terms, co-op fees, defect allowances, and compliance costs, the manufacturer may keep fifty to sixty dollars. The product has to be profitable at that number. If it is not, it cannot exist on the shelf regardless of how well it works.

Five tests. Not one of them asks whether this is the best product in the category. Not one asks whether the customer needs it. Not one asks what the customer would choose if they could see everything that exists.

Every requirement is rational from the retailer's seat. Slotting fees offset the cost of products that fail. Velocity thresholds prevent dead inventory. Payment terms finance the retailer's operations. The filter is not corrupt. It is efficient.

In 1967, Sol Price, the founder of FedMart, the discount chain that trained the man who would build Costco, wrote a memo to his employees. An investor named Nick Sleep would later frame the sentence on his office wall.14

Let us concentrate on how cheap we can bring things to the people, rather than how much the traffic will bear, and when the race is over Fed-Mart will be there. Sol Price, FedMart internal memo, 1967

The system Sol Price described was designed to serve the customer. Costco, built by his protégé Jim Sinegal, caps its markup at fourteen percent on outside brands and fifteen percent on its own Kirkland line.15 A typical grocery chain runs twenty-five to thirty. The low price is the reason over ninety-two percent of Costco members renew every year. The filter is how the system delivers that price.

The paradox is structural. The system that serves the customer also limits what the customer can see. Costco carries roughly four thousand products. Walmart carries over a hundred thousand.16

~4,000 Products at Costco
100,000+ Products at Walmart

The buyer at Costco decides, on behalf of every customer who walks through the door, which four thousand products will be available. What each of those customers actually needs is dispersed across millions of transactions, preferences, and local conditions that no single category review can access.17 The filter replaces that dispersed knowledge with five standardized questions. The questions are efficient. They are also a reduction. What does not fit the questions does not reach the shelf.

The same discipline that gives the customer the lowest prices in retail also gives the customer the fewest options. The savings and the absence come from the same source.

What survived

Three products the customer sees. They look normal, and all three survived the filter. The customer does not know the backstory. That is the design.

$1.89

The customer picks up a tall can of water for $1.89. It is branded like a beer. The company behind it, Liquid Death, sold three hundred and thirty-three million dollars of water last year and was valued at one point four billion.18 The water comes from mountain springs. It is not meaningfully different from what flows through a dozen other springs. The company is not yet profitable.19

What makes it work is the can, not the water. The branding is the product. The water is a commodity. To put that commodity in a hundred and thirty-three thousand retail doors, the company raised over two hundred and sixty-seven million dollars in venture capital.20 The money bought distribution, not product development. Without the fundraising, the can does not reach the shelf, the customer never sees it, and a different can of water sits in its place. Liquid Death did not survive the filter because the water is better. It survived because the capital was sufficient. The customer picks up the can. They are choosing the company that could afford to be there.

The can survived with branding and capital. The next product survived by cutting its margin in half.

Ninety-five percent

The customer buys a dog toy at Target for twelve dollars. They do not know that the company behind it went public in 2021 at a valuation of one point six billion dollars, or that the stock has since fallen ninety-five percent.21 They do not know that the same toy once sold as part of a subscription box at nearly twice the margin it earns in this store. DTC gross margins ran above sixty percent. In this store, about thirty-three.22

The company entered retail because the subscriptions slowed. Revenue has declined three years running: five hundred and thirty-five million, then four hundred and ninety, then four hundred and twenty-four.23 The company is shrinking while trying to stay on the shelf. The customer sees a dog toy at a price that looks normal. Behind the toy is a company that went public at twelve times the market cap it holds today, losing margin on every unit it sells in this channel, trying to survive in a format its economics were never built for.

BARK entered a shelf. The next company built its own.

A thousand stores

The customer walks into a Crumbl location. Pink box. Rotating flavors. The company scaled from one location to over a thousand in six years. The customer does not know what the franchise cost to open: between four hundred and sixty thousand and one point three million dollars, plus an eight percent royalty on every dollar of gross sales.24

They do not know that only forty-three percent of those locations are above average profitability, that the median owner netted two hundred and twenty-three thousand dollars last year on an investment that could exceed a million, or that roughly thirty-five stores have closed since the franchise started.25 The customer sees cookies. The storefront looks like success.

The missing

The customer sees what survived. What follows is what didn't.

385,000

Thomas Shaw invented a retractable safety syringe. The needle withdrew into the barrel after injection, making reuse and accidental sticks impossible. The Centers for Disease Control reported three hundred and eighty-five thousand needlestick injuries among hospital workers every year. Between sixty-two and eighty-eight percent were preventable with safer devices.26

When Shaw tried to sell the syringe in 1997, Becton Dickinson controlled ninety percent of the market and held sole-source contracts with the three largest group purchasing organizations, requiring member hospitals to buy ninety to ninety-five percent of their syringes from one supplier.27 Buying outside the contract meant losing bulk pricing across the entire supply catalog. Not just syringes. Everything. Bandages, gloves, tubing, sutures. The economics of the purchasing agreement made it irrational to buy the safer product, even when the cost of not buying it was measured in blood.

The safer needle existed. The purchasing system could not accommodate it. Healthcare workers developed HIV and hepatitis from injuries the retractable syringe was designed to prevent. Becton Dickinson eventually settled for a hundred million dollars. A jury awarded a hundred and thirteen million more in antitrust damages. Neither verdict undid the injuries.28

The syringe was blocked by a contract. The next product was blocked by a rebate.

$274

In 1999, a vial of Humalog insulin cost twenty-one dollars. By 2017, the list price was two hundred and seventy-four.29

$21 Humalog, 1999
$274 Humalog, 2017

An increase exceeding twelve hundred percent. The manufacturer released an authorized generic at half the price. Pharmacy benefit managers, the intermediaries that decide which drugs insurance will cover, excluded the cheaper version. The reason is structural: PBM rebates are calculated as a percentage of list price. A higher-priced drug generates a higher rebate. A cheaper version of the same molecule generates less revenue for the intermediary.

The three largest PBMs manage prescriptions for two hundred and seventy million Americans.30 In January 2025, the Federal Trade Commission reported that these three companies had generated seven point three billion dollars in markups on specialty generics over six years.31 The markups were not small. Tadalafil, a pulmonary hypertension drug, was marked up over seven thousand seven hundred percent. Dimethyl fumarate, a multiple sclerosis treatment, cost a hundred and seventy-seven dollars to acquire and was dispensed at close to four thousand.32 The cheaper drug existed. The patient did not receive it. By 2019, one in four insulin patients could not afford their medication.

The insulin was excluded by a financial incentive. The ice cream was excluded by a direct order.

"We need to shut this one down"

Jon Gordon was diagnosed with type 2 diabetes. He could not find a sugar-free ice cream worth eating, so he made one. Clemmy's sold in Ralph's and King Soopers. The product worked.33

Getting into more stores required slotting fees. Thirty thousand dollars for some Albertsons locations. Fifty thousand for some ShopRites. Nearly five hundred thousand for all Stop & Shop stores.34 In twenty-two of the twenty-five largest supermarket chains, Nestlé held the category captain role for frozen desserts, meaning the incumbent competitor advised the retailer on which products belonged in the freezer case.35 A 2007 Nestlé internal memo about Clemmy's read: "We need to shut this one down."36 In 2012, Nestlé told a Kroger frozen food manager to drop Clemmy's expansion or Nestlé would withhold distribution on its other ice cream products.37 The customer who was diabetic, the one who needed a sugar-free option, walked the freezer aisle and chose from what was there. Clemmy's declared bankruptcy in 2015.38

Clemmy's had a name. Most products that never reach the shelf do not.

The aisle

Rao's Homemade started as a restaurant brand in 1896 and became a small-batch pasta sauce. Getting it onto a national grocery shelf required two acquisitions. Sovos Brands bought it in 2017 to provide distribution infrastructure. Campbell Soup bought Sovos for two point seven billion dollars in 2024.39 Annie's Homegrown needed an eight-hundred-and-twenty-million-dollar acquisition by General Mills to reach the same shelf.40

The reader shops a grocery aisle every week. They see Prego, Ragú, Classico, and now Rao's. They do not see what else exists. They cannot. The products that were filtered out left no trace. No empty slot. No label that reads "removed." The absence is invisible by design.

What fills the space instead is the retailer's own brand. Kirkland Signature, Costco's private label, generated ninety billion dollars in revenue in 2025. More than Procter & Gamble's entire business.41

$90B Kirkland Signature, 2025
$84B Procter & Gamble, 2025

Private label products now account for twenty-one percent of every dollar spent in American stores, a record, and the share grows every year.42 The retailer's own brand pays no slotting fees, gets the best shelf placement, and answers to the same buyer who decides whether competitors stay or go. It is the ultimate survivor of its own filter.

The filter's benefit is visible. Lower prices. Reliable supply. Consistent inventory. What the filter removed is not.43

The paradox

Not all filters are the same. Sephora selects for innovation, digital presence, and brand story.44 Walmart selects for volume, price, and compliance. Costco selects for velocity and value, with roughly four thousand slots in the entire warehouse.45 Same customer, different store, different shelf, different absence.

Glossier launched in Sephora in February 2023 and sold over a hundred million dollars in its first year.46 Sephora compared the launch to Fenty Beauty and Rare Beauty. The same product would struggle at Walmart. Wrong price point, wrong volume requirement, wrong brand identity for a shelf next to Great Value. Sephora's filter selects for cultural heat. Walmart's selects for unit economics at forty-seven hundred stores. Different gatekeeper, different criteria, different shelf. The customer at each store sees the result. They do not see the filter that produced it.

The filter also changes over time. Products that survived last year's category review may not survive next year's. The shelf is not a permanent address. It is a lease that comes up for renewal at every review cycle.47 The customer sees the same brand in the same spot and assumes it has always been there. The company behind it knows the lease is conditional on velocity, margin, and compliance. One bad quarter and the space goes to whatever is next in line.

The system works. The filter produces low prices, reliable supply, and a shelf the customer can navigate in minutes. Without it, the store would be full of products that could not sustain themselves. The filter is how the system delivers value. It also produces invisible absence, and the system cannot deliver one without the other.

The shelf

The customer walks back into the store. Eight products. They pick one.

The shelf is a simplified version of a market too large and too complex to present in full.48 The customer reads the simplification and assumes it is the market. But a simplification, by definition, leaves things out. Everything filtered out is territory the customer never visits. The map looks complete because the territory it excludes has no representation. No blank space on the shelf that reads "something was here." No footnote that says "twelve others applied." The map is drawn, and the borders are invisible.

Every product on every shelf is the answer to a question nobody asks: why is this here and not something else? The answer is never "because it is the best." It is there because it survived the filter. The shelf selected for financial capability, not product quality. The customer chose from what made it through.

The products that were filtered out left no trace. No empty slot. No sign that reads "removed." The customer does not miss what they never knew existed.

Every product on the shelf would say the same thing. We got in.

New pieces when they're ready. Nothing else.

Sources

  1. Frederic Tudor, "The Ice King." First shipment: 130 tons of ice from Rockwood Pond, Saugus, Massachusetts, aboard the brig Favorite, departing Charlestown (Boston) on February 10, 1806, bound for Martinique. The voyage lost $4,500. Tudor spent time in debtor's prison in 1812-1813. The Calcutta route opened in 1833: the brig Tuscany carried 180 tons across 16,000 miles, crossing the equator twice. 100 tons survived the four-month voyage. Tudor built insulated ice houses in Havana, Calcutta, Madras, and Bombay. He packed ice in sawdust (free waste from lumber mills) and negotiated freight on ships sailing south with empty holds. Competitors who appeared were met with predatory pricing: Tudor dropped prices to 1 cent per pound until they withdrew. Gavin Weightman, The Frozen-Water Trade: A True Story (2003); Carl Seaburg and Stanley Paterson, The Ice King: Frederic Tudor and His Circle (2003); Smithsonian Magazine; New England Historical Society.
  2. In 1856, 363 ice-laden vessels left Boston Harbor carrying 146,000 tons. Ice was the second-largest U.S. export behind cotton. By 1880, U.S. ice consumption reached approximately 5.25 million tons. By the 1880s, ice delivery was widespread in major American cities. Henry David Thoreau, witnessing ice cutters at Walden Pond in the winter of 1846-47, wrote: "The pure Walden water is mingled with the sacred water of the Ganges." Thoreau, Walden (1854), Chapter 16; BU / The Brink; History Cambridge; Captains of Industry (1884).
  3. Approximately 30,000 new consumer packaged goods products enter American retail annually. Nielsen; Food Marketing Institute (FMI). The figure varies by year and category definition but has remained in the 25,000-35,000 range for over a decade.
  4. A Nielsen study of approximately 9,000 new items at a major U.S. retailer found roughly 60% were delisted within three years. Inez Blackburn (University of Toronto) estimates a 70-80% failure rate for new grocery products. Both figures consistent across multiple industry analyses.
  5. Slotting fees range from $25,000 per SKU in regional chains to $250,000 in high-demand metro markets. Per-store estimates: $250-$1,500 per SKU. Apple & Eve spent approximately $150,000 to place fruit punch in limited Safeway stores. Trax Retail; Pantry.ai; CBS News; Bay Food Brokerage.
  6. A national grocery chain launch for a single product requires an estimated $1.5 million to $2 million in slotting allowances alone, before accounting for free fill, promotional costs, or distributor margins. When a distributor manages slotting, the true cost can increase 50%+ due to added margins. Bay Food Brokerage; Bedrock Analytics; RangeMe.
  7. The FTC's most recent report on slotting allowances was published in 2003. No updated study, rulemaking, or enforcement action specifically targeting slotting fees has occurred in the two decades since. The practice remains legal and largely unregulated. FTC, "Slotting Allowances in the Retail Grocery Industry" (2003).
  8. Velocity thresholds vary by category. Refrigerated beverages: 5-7 units per store per week (USPW) to maintain shelf placement. Frozen meals: approximately 2 USPW. General rule: below 1 USPW signals risk of delisting. Buddy.io; Go Crisp; Balanced Business Group industry analyses.
  9. Retail deductions (chargebacks, markdowns, damages, promotional costs, compliance penalties) typically range from 5-15% of gross sales. Most brands average approximately 22% in combined trade spend and deductions off gross revenue. Disputed or unrecoverable write-offs account for an estimated 10-20% of all deductions. Woodridge Retail Group; Rodeo CPG; Inmar Intelligence.
  10. Jordan Buckner, founder of FoodBevy (kombucha brand), invoiced national distributor KeHE for $32,000 and received $21,000. The $11,000 gap (34% of gross revenue) comprised chargebacks, fees, and returned products. The largest single charge ($8,400) came from Jewel-Osco returning unsold inventory. Buckner stopped selling through KeHE to avoid insolvency. His recommendation: stay with regional distributors until $5M+ in annual revenue. Booch News, "The Hidden Costs of National Distribution" (October 2024).
  11. Target Perfect Order Program compliance requirements. Barcode accuracy penalty: $0.75 per carton with $100 minimum for unscannable barcodes. Target Partners Online vendor documentation; industry compliance guides.
  12. Costco requires a minimum of $5 million in commercial general liability insurance per occurrence for all vendors. Binary gate: no insurance, no entry. Vividly, "Working With Costco: A Guide for CPG Brands"; industry vendor compliance documentation.
  13. Major big-box retailers typically command 35-50% of retail price on general merchandise. Grocery margins run lower (25-30%). Private label products yield 25-35% gross margin, typically 25-30% higher than comparable national brands. NielsenIQ; Private Label Manufacturers Association (PLMA); RetailWire.
  14. Sol Price, FedMart internal memo, 1967. Quoted by Nick Sleep in the Nomad Investment Partnership Interim Letter, June 2010. Sleep described the memo as framed on his office wall and called it "the best summary of the business case for scale economics shared we have come across." Full text: "Although we are all interested in margin, it must never be done at the expense of our philosophy. Increasing the retail prices and justifying it on the basis that we are still 'competitive' could lead to a rude awakening as it has with so many." Sol Price founded FedMart (1954) and Price Club (1976), which merged with Costco in 1993. Source traced to Robert Price, Sol Price: Retail Revolutionary & Social Innovator (2012).
  15. Costco markup caps: 14% on third-party products, 15% on Kirkland Signature. Blended gross margin approximately 12-13%. Membership renewal rate: approximately 92.3% in US and Canada (fiscal 2025). Blended gross margin approximately 10.9-11%. Costco investor relations; Acquired.fm (Costco episode); Motley Fool.
  16. Costco carries approximately 3,700-4,000 active SKUs per warehouse. Walmart carries over 100,000. A typical grocery store carries approximately 30,000. Tinuiti Costco Vendor Guide; Produce News; Supply Chain Game Changer.
  17. The impossibility of centralizing dispersed knowledge into standardized selection criteria draws on Friedrich Hayek, "The Use of Knowledge in Society," American Economic Review (1945). Hayek's central observation: "The shipper who earns a living by knowing where empty or half-full cargo space is available, the real estate agent who knows local conditions" possess knowledge that "cannot be communicated to a central authority because it is ephemeral, contextual, and often tacit."
  18. Liquid Death revenue: $263 million (2023), $333 million (2024, +27% year-over-year). Valuation: $1.4 billion (Series E, March 2024). The company hired Goldman Sachs in July 2023 to explore an IPO; no S-1 has been filed. Sacra; Retail Dive; Fueler; Access IPOs.
  19. Liquid Death has stated it is "not yet profitable" but "making big strides to get there." Water sourced from U.S. mountain springs, bottled in Bland, Virginia and Mackay, Idaho. Originally sourced from Austrian Alps. Not municipal or tap water. Sacra; Honest Brand Reviews; Wikipedia.
  20. Liquid Death has raised over $267 million in total venture capital funding across multiple rounds. Series E (March 2024): $67 million at $1.4 billion valuation. Prior rounds: Series D ($70M, October 2022 at $700M valuation), Series C ($75M, 2022). Retail Dive; Sacra; Prime Unicorn Index.
  21. Liquid Death retail distribution: 133,000+ retail doors (2024), up from approximately 16,000 in 2021 (731% store growth over three years). Retailers include Walmart, Target, Kroger, Albertson's, Whole Foods, 7-Eleven, Circle K. TapTwice Digital.
  22. BARK (BarkBox) merged with Northern Star Acquisition Corp via SPAC in June 2021, valued at $1.6 billion. All-time stock high: approximately $19.54 (December 2020). Price as of March 2026: approximately $0.81, a decline of roughly 95-96%. Market capitalization approximately $135 million. Retail Dive; Pet Food Processing; MacroTrends; Morningstar.
  23. BARK DTC gross margin: approximately 60.9% (Q2 FY2023), rising to 67% record (Q1 FY2026). Commerce/Retail gross margin: approximately 33.2% (Q2 FY2023). DTC represents 86-89% of revenue. BARK Q2 FY2023 press release; BARK FY2024 and FY2025 investor results.
  24. BARK revenue: $535.3 million (FY2023), $490.2 million (FY2024, -8.4%), approximately $424 million (trailing twelve months, December 2025). Revenue declining three consecutive fiscal years. BARK FY2024 and FY2025 investor results; BusinessWire.
  25. Crumbl franchise startup cost: $460,166 to $1,266,333 (per FDD; other sources cite up to $1,442,533). Franchise fee: $50,000. Royalty: 8% of gross sales. Marketing fee: 2% of gross revenue. Franchise Direct; Franchise Chatter (FDD data).
  26. Crumbl franchise data from Franchise Disclosure Document (FDD). 2024 data for reporting locations: only 43% of 1,000+ locations above average profitability. Median net profit: $223,236 (2024, per revised FDD; initial filing reported $77,359 using a single-store proxy, later corrected). Average unit volume: approximately $1.35 million. Total closures: approximately 35-37 across all years (7 in 2023, 12-14 in 2024, 16 in 2025). Total locations: approximately 1,103 US + 25 Canada as of early 2026. Restaurant Business Online; Franchise Chatter; Franchise Times; Inc.
  27. CDC NIOSH Alert 2000-108: estimated 385,000 percutaneous injuries from contaminated sharps among hospital-based healthcare workers annually. CDC estimates 62-88% of sharps injuries are preventable with safer needle devices. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health.
  28. Retractable Technologies case. Thomas Shaw invented the retractable safety syringe. Becton Dickinson controlled approximately 90% of the U.S. syringe market. Group Purchasing Organizations (Novation, Premier, HealthTrust) required member hospitals to purchase 90-95% of syringes from contracted suppliers to receive maximum discounts. Buying outside the contract resulted in loss of bulk pricing across all supply categories. Senate Judiciary Committee testimony, Joe Kiani, April 2002; Senate hearing CHRG-107shrg85986; BioSpace.
  29. Healthcare workers developed HIV and hepatitis from needlestick injuries that retractable syringes were designed to prevent. Novation, Premier, and Tyco settled in April 2003. Becton Dickinson settled for $100 million in July 2004. A 2013 jury awarded $113.5 million in antitrust damages (later reversed on appeal). Senate hearing CHRG-107shrg85986; Blurred Bylines; CDC NIOSH.
  30. Humalog (insulin lispro, Eli Lilly): list price rose from $21 in 1999 to over $274 by 2017, an increase exceeding 1,200%. Eli Lilly introduced an authorized generic at 50% lower list price. PBMs systematically excluded the cheaper version from formularies because branded insulins generated higher percentage-based rebates. By 2019, one in four insulin patients could not afford their medication. US Pharmacist; FTC press release, September 2024.
  31. Top three PBMs (Express Scripts, CVS Caremark, OptumRx) control approximately 80% of all prescription claims, managing prescriptions for approximately 270 million Americans. Market concentration index (HHI): 1,972 (threshold for "highly concentrated" is 1,800). IntuitionLabs; Kaiser Family Foundation (KFF).
  32. FTC Second Interim Staff Report (January 2025): the Big 3 PBMs generated $7.3 billion in excess dispensing revenue from specialty generic markups at affiliated pharmacies between 2017 and 2022. In February 2026, the FTC secured a settlement with Express Scripts requiring net-cost-based pricing. FTC; Fierce Healthcare; CNN.
  33. Individual drug markups documented in FTC Second Interim Staff Report: tadalafil (pulmonary hypertension) marked up over 7,700% in 2022. Dimethyl fumarate (multiple sclerosis) acquired at $177 and dispensed at close to $4,000 for a 30-day supply. Top 10 specialty generics alone generated $6.2 billion in excess dispensing revenue. FTC Second Interim Staff Report (January 2025).
  34. Jon Gordon founded Clemmy's after a type 2 diabetes diagnosis. The product sold successfully in Ralph's (California) and King Soopers (Colorado) in 2010-2011. Food Navigator-USA, "Pay-to-play slotting deals penalize small players" (August 2014).
  35. Clemmy's slotting fee amounts: $30,000 for some Albertsons stores; $50,000 for some ShopRite stores; nearly $500,000 would have been required for all Stop & Shop locations. Jon Gordon's quote: "Between slotting fees and the category captain system, those two things alone make it almost impossible for a smaller manufacturer to make it." Food Navigator-USA; CBS News.
  36. Nestlé held the category captain role for frozen desserts in 22 of the 25 largest U.S. supermarket chains. The category captain advises the retailer on assortment, placement, and pricing, while simultaneously competing against the brands it evaluates. CBS News, "How Grocers Wring Extra Cash Out of Shelf Space" (2016).
  37. "We need to shut this one down." Internal Nestlé memo, 2007, regarding Clemmy's. Cited in Food Navigator-USA (August 2014), referencing court documents and legal proceedings.
  38. In early 2012, Nestlé allegedly told Kroger's lead frozen food manager to drop Clemmy's expansion, reduce its SKUs, and significantly increase its unit price, or Nestlé would not contract with Kroger to distribute Arctic Zero and other Nestlé ice cream products. Food Navigator-USA (August 2014), citing allegations in legal proceedings.
  39. Clemmy's declared bankruptcy in 2015. Food Navigator-USA; CBS News.
  40. Rao's Homemade: Sovos Brands acquired the brand in 2017 to provide distribution infrastructure. Campbell Soup acquired Sovos Brands for $2.7 billion in 2024. Under Campbell's distribution, Rao's reached Walmart, Kroger, Albertsons/Safeway, Whole Foods, and Sprouts nationally. Yahoo Finance; Food Dive.
  41. Annie's Homegrown (founded by Annie Withey) was acquired by General Mills for $820 million in 2014 to achieve national distribution. The pattern repeats across small brands: Late July Snacks founder Nicole Bernard Dawes described the same dynamic: "When I made my first sales call for tortilla chips in 2010, I desperately needed to land a retailer to also secure a distributor." Late July was later acquired by Campbell's via Snyder's-Lance in 2018. Food Dive; General Mills investor relations; Fortune (March 2026).
  42. Kirkland Signature (Costco's private label) generated approximately $90 billion in revenue in fiscal year 2025 (ended August 31, 2025). Larger than Procter & Gamble's total revenue (~$84.3 billion, FY2025). P&G Annual Report 2025. Accounts for roughly 28-33% of Costco's total merchandise sales. Motley Fool (March 2026); Chowhound.
  43. U.S. private label sales reached $282.8 billion in 2025, a record. Dollar market share: 21.3% (all-time high). Unit market share: 23.5% (all-time high). Five-year growth: dollar sales up $64.8 billion (+30%) since 2021. Dollar share rose from 19.1% to 21.3% over the same period. Circana Unify+ data (52 weeks ending December 28, 2025); PLMA; Supermarket News.
  44. The distinction between what the customer sees (the shelf) and what the customer cannot see (what was filtered out) draws on Frédéric Bastiat, "Ce qu'on voit et ce qu'on ne voit pas" (1850). Bastiat's central observation, paraphrased: the seen benefit is always concentrated and visible; the unseen cost is always dispersed and silent. This formulation synthesizes Bastiat with Mancur Olson's concentrated benefits / diffuse costs framework. Bastiat's original text: "In the department of economy, an act, a habit, an institution, a law, gives birth not only to an effect, but to a series of effects. Of these effects, the first only is immediate... the others unfold in succession." The customer sees the result of the filter. The cost, the products that were removed, has no advocate, no label, and no presence.
  45. Sephora prioritizes brands with "unique, innovative products that stand out in the market with a high digital footprint." Selection criteria include brand story, packaging, sustainability practices, and trend alignment. Sephora runs an Accelerate program for emerging brands. Royal Labs; Inside Sephora.
  46. Costco carries approximately 3,700 active SKUs per warehouse. Products are selected for velocity, member value, and bulk efficiency. All goods ship, replenish, and fulfill in full pallet quantities. Maximum gross profit margin: 14% (15% for Kirkland Signature). Costco will launch Kirkland Signature versions when quality and demand justify it. Tinuiti Costco Vendor Guide; Produce News; Supply Chain Game Changer.
  47. Glossier launched in Sephora in February 2023 across approximately 600 stores in the U.S. and Canada. Sephora sales exceeded $100 million in the first full year, "exceeding Sephora's forecasts by over 100%." Sephora compared the launch to Rare Beauty and Fenty Beauty. Total 2023 retail sales approximately $275 million (+73% year-over-year). The company reached profitability under CEO Kyle Leahy. BeautyMatter; Fast Company; WWD; Glossy.
  48. Products that survive the initial shelf placement face ongoing category reviews, typically conducted quarterly or biannually. Products that fall below velocity benchmarks, lose margin share to private label, or fail compliance audits risk discontinuation at each review. The shelf is conditional, not permanent. Industry standard practice across major retailers.
  49. The framework of reading the shelf as a simplified map of a larger, more complex territory draws on James C. Scott, Seeing Like a State (1998). Scott's central observation: the act of simplifying a complex system into a legible, governable form destroys the local knowledge ("metis") that made the original system work. The epigraph from Brian Friel's play Translations: "OWEN: We're making a six-inch map of the country. Is there something sinister in that? YOLLAND: Something is being eroded." The retail shelf is a legibility project. Only products that are legible to the system, those with UPC codes, EDI capability, compliant packaging, sufficient insurance, and demonstrable velocity, pass through. Products that are excellent but illegible are invisible.