Natalie Holloway started Bala with $5,000 and shipped 75,000 units of wearable weights from her Hollywood garage while working full-time. The product resonated. Bloomingdale's carried it. Free People carried it. Orders kept growing.

Then the pandemic fitness boom hit. Revenue went from $2 million to $20 million in a single year. The product hadn't changed. The market had found it.

Within months, the same founder who had scaled 10x was laying off 90% of her team and coming within inches of bankruptcy. The fitness industry reversed. She had hired and invested based on the growth rate. The infrastructure couldn't absorb the reversal.

The product was the same product that won the growth. What failed was the structure around it. The forecasting, the inventory commitment, the staffing model, the assumption that the trajectory would hold. Every one of those decisions was made inside a system that rewards growth and has no mechanism to signal when the growth is structural versus temporary.

Products don't fail on shelf. Systems filter them out.

What the failure rate actually measures

Nielsen puts it at 85% for new CPG products. Research at a major U.S. retailer found that only about 40% of new items were still being sold three years later. Inez Blackburn at the University of Toronto estimated 70-80% failure in grocery specifically. Parade, a DTC underwear brand that generated $21 million in revenue in 2022, burned through it and folded entirely by October 2025.

The standard explanation is product quality. The product didn't resonate. The market wasn't ready. The brand didn't find product-market fit.

But Bala had a product that sold. Kendra Scott had a product that department stores loved. Parade had revenue in the tens of millions. The products were not the variable that changed. The structure around them was.

Five participants, five metrics, zero alignment

The retail supply chain has at least five participants making decisions simultaneously. Each one optimizes a different metric.

The retailer optimizes margin per item and sell-through velocity. Products that don't move get discontinued. The buyer is not evaluated on whether your brand succeeds long-term. They are evaluated on whether the shelf performs this quarter.

The brand optimizes gross margin on the wholesale price. As piece one in this series showed, that model is missing half the inputs.

The marketing agency optimizes ROAS and CAC on digital campaigns. As piece two showed, those metrics cannot see retail sell-through.

"We started to get a few returns. Two returns. Five returns. Twenty returns. Fifty returns. One hundred returns. And then the stores were like, we're sending all of it back."

Keon Gazari, sourcing expert

The manufacturer optimizes unit cost. Keon Gazari, a sourcing expert who has placed over 2,500 products, described what happens when this optimization runs unchecked. A military boot manufacturer reduced the rubber content in its soles from 40% to 32% across two full shipping containers. The manufacturer saved a fraction of a cent per sole. The brand absorbed the full return.

The consumer optimizes value for money. They don't know or care about any of the above.

Five participants. Five metrics. Each rational. Nobody optimizing the outcome that determines whether the brand survives.

What "getting on shelf" actually selects for

The buyer meeting is the most misunderstood conversation in retail. Most founders prepare to answer "why is your product good?" The buyer is not asking that question.

The buyer is evaluating three things. First, reorder likelihood. The buyer has limited shelf space. Every slot occupied by your product is a slot not occupied by a proven performer. The buyer wants evidence of demand, not evidence of quality. Second, logistics competence. The routing guide, the EDI integration, the labeling requirements, the ship windows. Product quality does not predict operational execution. Third, margin durability. The wholesale price, the promotional requirements, the markdown risk. A beautiful product priced wrong for the category kills the deal regardless of quality.

"It's not about why your product is better than what's on the shelves. It's more about how are you growing that category as a whole."

Jaime Schmidt, founder, Schmidt's Naturals

Jaime Schmidt described learning this at Costco, Target, and Walmart. The buyer's incentive is category growth. A brand that grows the category gets shelf space. A brand that tells a beautiful story but doesn't move the category metric gets discontinued.

The shelf is not a destination. It is an audition the brand is paying for.

The sequential dependency problem

Getting on shelf is one event. Selling through is a second event. Getting the reorder is a third. Each depends on the one before it. Most brands optimize everything for the first event and arrive at the second with no plan.

The brand spends months preparing the buyer pitch. The packaging is perfect. The sell sheet is polished. The brand gets the purchase order. Then the product is on shelf and the promotional calendar starts. The sell-through data trickles in. The marketing team is optimizing a dashboard that can't see in-store performance. The retail deductions start arriving.

By the time the brand has enough data to understand what's happening, the buyer is already reviewing the category for next quarter. If velocity is below the threshold, the brand is flagged.

Holly Thaggard, the founder of Supergoop, made a structural move most founders never get to make. Sephora was treating her SPF brand as seasonal and planned to pull it off shelf in September. Instead of fighting for more space, she negotiated for less space on a permanent commitment. Six inches of shelf, 365 days a year. She traded shelf width for shelf permanence, which is the trade most brands need but never see in time. By the time most founders recognize the discontinuation pattern, the review is already scheduled.

Why the failure rate is structural

The system was designed to maximize the retailer's option value. Put many products on shelf. See what moves. Discontinue the rest. Repeat.

This is rational for the retailer. The cost of a failed product is low. Markdown money and return terms push the financial risk onto the brand. The data from the failed product informs the next assortment decision. The shelf slot is refilled within a quarter.

The cost for the brand is existential. The inventory is manufactured. The cash is deployed. The promotional commitments are made. When the product is discontinued, the brand absorbs the write-down, the dead inventory, and the lost promotional spend.

Kendra Scott built her jewelry business entirely on wholesale. By 2008, department stores were her primary channel. Then the recession hit. Buyers were laid off. Boutiques shuttered. Her bank called and asked her to pay off her line of credit within six months. She had never been late on a payment. The system processed her anyway.

The question the failure rate is really answering

The routing guides were designed for P&G. The promotional calendars were designed for companies with trade marketing departments. The margin architecture was designed for businesses that can absorb chargebacks across hundreds of SKUs. The metrics were designed for a single-channel world.

A brand with eight employees entering this system for the first time is not failing because the product is wrong. It is failing because the institution was designed around a company it is not.

The product might be excellent. The founder might be talented. The market might be ready. None of that changes the structure the brand is entering.

We have watched brands walk into this. The product was good. The team was capable. The market was ready. The structure processed them anyway. Most assume they did something wrong. They didn't.

The failure rate is not a verdict on the product. It is a measure of the distance between what the founder was told retail would be and what retail actually is.

MMK Retail helps consumer brands see the design of the retail institution before the institution shows itself.