What game are you in

Recent Marketplace Pulse article reported that, as of early 2026, just 1.6% of active sellers (fewer than 8,000 operators) are responsible for roughly 50% of Amazon’s U.S. third-party GMV, which is estimated at around $300 billion.

Three years ago, in 2023, it took around 15,000 sellers to account for that same share of GMV. In under three years, the threshold of dominance has nearly halved. That’s quite a remarkable, and not in a good way, level of compression.

The concentration becomes even more striking at the upper tiers: a group of 111 sellers generates 10% of third-party GMV. Roughly 1,020 sellers generate 25%.

These figures by the way resemble broader wealth concentration patterns in the United States, where the top 10% of households hold roughly 67–70% of total wealth. Similar direction: a shrinking share of participants account for a growing share of outcomes.

Marketplace Pulse frames this development as part of what it calls the “Great Compression”, a period characterized by fee increases, rising advertising costs, growing algorithmic complexity, and higher operational demands. The performance bar has risen across multiple dimensions at once.

Whenever the bar rises, the question that all participants are forced to think through is: what game are we actually playing? Because when performance and expectations are higher, you have to look for leverage, unfair advantage to thrive when everybody else is just as good and works just as hard.

So, let’s define what is middle class on Amazon. When I use the term “middle class” on Amazon, I am referring to sellers who are competent but not structurally advantaged. They have good products, good listings, and functional advertising programs. They are not negligent operators. Yet they lack a moat.

And moats are difficult to build.

In consumer goods, the most durable moats rarely originate on Amazon itself. They are innate to the brand and either executed within Amazon, or amplified by it.

Product is the most obvious example. Many brands have good products. Very few have exceptional ones. If a brand has a true virality of its products (aka you love it so much you have to tell all your friends, vs. a TikTok influencer virality), other building blocks go much faster. Mineral-based sunscreen that spreads and feels identical to conventional chemical sunscreen? A high-protein, low-sugar granola bar that delivers the indulgent experience of a pop tart? (examples from my life). Unimaginably hard. In “better-for-you” categories, sensory experience is extraordinarily difficult to optimize. That difficulty is precisely why true product superiority is defensible.

Intellectual property can reinforce this moat, particularly in categories where formulation or functional innovation is meaningful. IP alone does not guarantee success, but when combined with demonstrable superiority, it gives structural insulation.

For smaller and emerging brands, community has emerged as a modern moat. Liquid Death is by now a classic example of building a brand in a commoditized category on identity-driven purchase. It was quick for them to blow up on Amazon because they did hard work before that. Community is upstream of Amazon, but it changes Amazon performance because branded search increases, conversion improves, reviews grow, and all of that helps paid marketing perform.

Virality and narrative can also create leverage, though they are harder to sustain without underlying product strength.

Another moat I increasingly see among stronger operators is speed of learning, production, and iteration, particularly in two areas: SKU portfolio development and marketing. With AI automation across Meta, Google, and increasingly Amazon, the current new world rewards the rapid deployment of creatives so the algorithm can optimize bidding and placement. In the SKU portfolio scale game the advantage is the speed of a professional operator: launching quickly, going deeper within a category, building cross-sell logic, protecting margins, and decisively killing underperforming SKUs. This advantage of speed is less visible and sexy than product innovation or community building, but there are many profitable 8+ figure ecommerce brands that run on this.

When I look at a business I try to identify its superpower. And that superpower is rarely an Amazon tactic. It is something innate to the brand that Amazon can amplify.

ANd data on concentration of top Amazon sellers isn’t meant to be discouraging, but create more clarity. If the middle is shrinking, then average positioning and moderate execution are unlikely to produce durable outcomes. Question I keep coming back to - for myself and for the brands I work with - what game are we playing? Amazon rewards outsized advantage, finding, and being able to stick to building that advantage is the real work.

Saludos,

Irina