The Beliefs That Kill Products

By most estimates, 70% to 80% of new consumer products fail within two years of hitting a retail shelf.¹ That number has been published, cited, and ignored for decades. It keeps not changing.

The standard explanation is that these products were bad. Wrong features. Wrong timing. Wrong market. That explanation is comfortable because it implies the failure was visible. That someone should have known. That yours will be different because your product is better.

The actual pattern is less forgiving. After 25 years inside retail systems and over a thousand product programs, the failures I've seen share a different root cause. The products weren't bad. The economics were wrong. And the economics were wrong because the founders were operating on inherited beliefs they never verified.

Three beliefs in particular. Each one is widely held. Each one is structurally incorrect. And each one is responsible for more product deaths than any competitor, any market shift, or any bad review.

Belief #1: "My margins work."

Ask a founder what their margin is and they'll give you a number. It's almost always the same calculation: retail price minus cost of goods, divided by retail price. A product that costs $150 to make and sells for $499 has a 70% margin. The math is clean. The math is also wrong. I've watched founders present this number to retail buyers who already know it won't survive their deduction schedule. The buyer never corrects it.

That 70% is gross margin on paper. It has nothing to do with what ends up in your bank account. Between the factory and the deposit, at least 11 categories of cost eat into that number. Most founders undercount them or miss them entirely:

For a DTC brand, customer acquisition alone can run 30-40% of revenue. That one line item cuts a "70% margin" product to 30 or 40 percent before anything else is counted.

But the cost chain is only half the problem. The other half is what the retailer takes.

Every major retailer takes a margin. The ranges vary by channel and category, and most founders encounter them for the first time after the purchase order is signed:

Then come the deductions. Promotional allowances. Volume rebates. Advertising contributions. Damage claims. Chargebacks for labeling errors, late shipments, or pallet configuration mistakes. Trade promotions alone consume about 20% of revenue for CPG manufacturers, one of the largest lines on the P&L.² Roughly 60% of those promotions never break even.³

Then there's cash. Canadian Tire pays Net 90. That means a founder funds three months of inventory, freight, warehousing, and overhead before seeing a single dollar of revenue. For a product with $205 in true landed cost shipping 200 units a month, that's $123,000 in cash you need to have before you start.

What this looks like on a real product

Take a product with $150 in factory cost, selling at $499 online and $579 in store. Run it through three channels. Same product. Same COGS. Radically different economics:

The DTC channel looks profitable at $160 per unit. But the true landed cost is $338.82, primarily because the brand is paying for its own customer acquisition, fulfillment, and returns. The DTC margin evaporates if monthly volume drops below 160 units. The margin looks healthy. The business model is fragile.

Costco looks like the best economics. And it is, per unit. But it requires $79,000 in working capital before the first reorder. And Costco's velocity threshold means that if the product doesn't move at roughly $1,500 per warehouse per week, it is on a clock. Below about $500, it gets deleted. Fast.⁴

Canadian Tire has the worst combination: a 38% retailer margin, the highest working capital requirement ($123,000), and Net 90 payment terms. A founder who celebrates getting into Canadian Tire may be celebrating the beginning of a cash crisis.

The first belief, "my margins work," fails because founders are comparing retail price to factory cost. The actual comparison is retail price minus retailer margin minus deductions minus 11 categories of cost-to-serve, adjusted for payment terms and working capital. Most founders have never seen this math. They discover it one line item at a time, over months, as each cost reveals itself.

Belief #2: "Getting on shelf is winning."

A purchase order from a major retailer feels like validation. It is not. It is a starting position.

Every retailer has a velocity threshold. A minimum sales rate the product must sustain to keep its shelf space. Fall below it and the product gets delisted. These numbers are not published. They are not negotiable. Most founders find out they exist when the delisting notice arrives.

At Costco, the math is binary. A $499 product needs to sell roughly 3 units per warehouse per week to clear the base threshold. Below that, the product is on a clock. Costco's inventory turns at 13x per year, among the fastest of any major retailer. Products that don't move are deleted fast.

This mechanism is not linear. It's binary. Above threshold, the flywheel works: reorders come, shelf space holds, the brand builds history with the buyer. Below threshold, value destruction begins. I've watched products go from first placement to delisted in under ninety days. Across 16 validated cases, the marginal zone (0.85x to 1.2x of threshold) usually resolves downward within one to three years.

The velocity decay problem

A product's launch velocity is almost always its peak. Without organic repeat demand, every reorder cycle loses volume. How fast it decays depends on whether the product generates its own repurchase:

A product graded B+ at launch starts at 1.00 and decays to 0.88 by the third reorder cycle. If the launch velocity was 1.1x of threshold, the product crosses below threshold by cycle 3. The founder thinks they have a year. They have months.

Solo Stove: the Snoop Dogg proof case

In 2023, Solo Brands spent $96.9 million on advertising, a year built around a Snoop Dogg campaign for Solo Stove.⁹ The campaign generated massive brand awareness. It did not change the binary mechanism. In January 2024 the company's interim CFO said the campaigns raised awareness but "did not lead to the sales lift that we had planned."¹⁰ The campaign moved awareness. It did not move the fundamental math. The stock lost more than 99% of its value.

Awareness and velocity are not the same thing. A product can be famous and still fail the binary test. The shelf doesn't care how many people know your name. It cares how many people buy your product this week, at this store, at this price.

The second belief, "getting on shelf is winning," fails because placement is not the outcome. The reorder is the outcome. A product that gets placed and doesn't hit velocity threshold has not succeeded. It has created a track record of poor performance that follows the brand to every future buyer meeting.

Belief #3: "My product is different."

This is the most dangerous belief because it's partially true. The product might genuinely be different. Better engineered. Better designed. Solving a real problem in a new way. None of that changes the structural ceiling.

Two factors set the ceiling. No amount of product quality changes either one.

The first is whether the category repeats. Coffee and cleaning supplies bring customers back in days or weeks. A grill, a generator, a stand mixer does not. The category's repeat cycle sets a hard ceiling on reorder velocity. No amount of product quality moves it.

The second is fan-channel fit. Do the people most likely to buy this product actually shop at this retailer? A premium outdoor product at Costco has strong fit. Costco's membership skews toward homeowners with disposable income who already buy in outdoor and patio. The same product at Dollar General has near-zero fit regardless of price.

Both factors are structural. Product quality, marketing spend, pricing, operational execution. None of it moves them. A product can score perfectly on every performance dimension and still be capped by its structure.

The Instant Pot pattern

Instant Pot became one of the most recognizable kitchen brands in North America. It dominated Amazon Prime Day. It had genuine word-of-mouth growth and a passionate community. In 2023, Instant Brands filed for bankruptcy.

The structural problem: Instant Pot was a durable product in a non-repeating category with wave-driven demand. The initial adoption wave created explosive growth. When the wave broke, there was no repeat cycle to sustain velocity. Every household that wanted an Instant Pot had one. The reorder disappeared. The shelf space followed.

Compare this to KitchenAid. Similar price point. Similar category (kitchen appliances). But KitchenAid solved the structural ceiling problem with an ecosystem: dozens of attachments, most priced between $30 and $200. The stand mixer is the anchor. The attachments are the repeat. KitchenAid leads its category and sustains reorders at every major retailer it operates in.

Or YETI. The cooler ($300+) is the anchor. The tumblers ($35) are the repeat. YETI's structural grade is A because the ecosystem generates quarterly repurchase in a category that would otherwise be a one-time buy.

The question for every durable product: What is the $29-99 product that brings customers back every quarter? If the answer is "nothing," the structural ceiling caps the product at a B+ in the best case, regardless of every other factor.

The third belief, "my product is different," fails because different is a performance claim. The ceiling is structural. A product can be genuinely superior on every axis that matters to the consumer and still be unsustainable at retail because the category doesn't repeat and the ecosystem doesn't exist.

The pattern

The three beliefs run the same way.

1. The founder holds a belief that appears reasonable on the surface.
2. The belief is inherited from industry convention, not verified against the specific economics of their product at their retailer.
3. The belief fails silently. There is no single moment of failure. The margin erodes gradually. The velocity decays over cycles. The structural ceiling reveals itself only after the reorder doesn't come.
4. By the time the failure is visible, the capital is spent, the shelf space is lost, and the track record is created.

This is not a knowledge problem. Retailer margins are discoverable. Velocity thresholds can be estimated. Repeat cycles can be assessed. The information exists. Anyone willing to look can find it.

It is a verification problem. The beliefs persist because nobody forces the question before the money is spent. The purchase order arrives, the celebration happens, and the beliefs ride along unchallenged into the supply chain, the cash flow projections, and the board presentations.

I know this because I've done it myself. Trusted the premise. Accepted the inherited math. Did not verify the economics at the level of specificity required to see the failure coming. It cost me real money and two years I don't get back. The details are mine. The pattern is universal.

What we built

The response to this pattern was not a framework, a course, or a set of principles. It was a discipline. A sequence of questions that must be answered, in order, before any capital gets committed. Each question exists because I watched someone skip it and pay for the omission.

The first question that gets skipped is whether demand is real or manufactured. A product with strong DTC sales driven by paid acquisition looks identical to a product with genuine product-market fit. The economics are completely different. Most founders have never been asked to distinguish the two.

The second question is whether the unit math survives contact with the retailer's actual cost structure. Not gross margin. True landed cost, through the full deduction chain, at a specific retailer, with their margin structure, payment terms, and reserve requirements applied. The math in this article is not theoretical. It is the math most founders encounter for the first time after the purchase order is signed.

The third question is structural. Will the product sell fast enough to sustain its shelf space? This is where most founders want to start. It is the wrong place to start. Structure comes before performance. Category repeat cycle and fan-channel fit set the ceiling. Everything else operates underneath it.

The fourth question is operational. Can the brand execute at this retailer's standard? Barcodes, packaging specifications, pallet configurations, fill rate requirements, EDI compliance, insurance, safety certifications. Every retailer has different requirements. A single compliance failure generates chargebacks that can erase the margin the product was supposed to make.

The fifth question works backward from the retailer's velocity threshold. Given projected sales curves, marketing assumptions, and organic demand estimates, will this product hold its shelf space past the third reorder? The answer resolves to one of three verdicts: sustains, fragile, or will not hold.

The sequence matters. A product whose unit math does not work never reaches the velocity question. A product whose math works but whose structure will not support reorders gets a clear signal early: the economics are sound, but the shelf will not hold. Adjust the strategy or do not enter.

Every question in this sequence was born from a specific failure. A specific product, at a specific retailer, where a specific belief went unchecked and a specific amount of money disappeared. The discipline is scar tissue turned into structure. It works because it makes it impossible to skip the uncomfortable question.

Implication

The math is not the point. The discipline is.

A founder can run every calculation in a spreadsheet. The math is not proprietary. Retailer margins are discoverable. Velocity thresholds can be estimated. Cost-to-serve can be mapped. None of this requires specialized software.

What requires infrastructure is the forcing function. The mechanism that ensures the question gets asked before the purchase order gets signed. Without that mechanism, the beliefs ride along unchecked because the incentives favor optimism. The founder wants to believe the margins work. The broker wants to believe the placement will sell. The retailer wants to believe the product will clear threshold. Everyone benefits from not asking the question until it's too late to change the answer.

Seventy to eighty percent failure rates do not persist for decades because the products are bad. They persist because the verification step doesn't exist in the standard workflow. The beliefs get inherited, the celebration happens, and the math reveals itself one chargeback at a time.

The question is not whether your product is good enough. The question is whether you've checked the three beliefs that kill products before you've spent the money that proves them wrong.