Consumer brands win at 10 to 15 percent better, not 10x better
Silicon Valley doctrine demands a 10x improvement before a product deserves to exist. Allison Ellsworth, who built Poppi into a brand that landed a Shark Tank deal, argues that threshold is wrong for consumer goods, and her reasoning cuts closer to how consumer behavior actually works.
The 10x improvement doctrine sits near the center of Silicon Valley’s founding mythology. Build something an order of magnitude better, or do not bother. Allison Ellsworth, who co-founded the beverage company Poppi, argues the doctrine is wrong, at least for consumer products. “The best brands and the best businesses are 10 to 15% better than the established norm,” she says. That is a direct rebuttal of a principle many founders treat as settled.
The logic behind Ellsworth’s position is not contrarian for its own sake. Consumer behavior is shaped by habit, familiarity, and existing emotional attachments to products. A radical departure does not tap into those things; it asks the consumer to abandon them. An incremental improvement, precisely targeted, can travel along the grain of existing preference rather than against it. Poppi’s own path bears this out. The brand took soda, a product with deep nostalgic roots, and improved it on two specific dimensions: less sugar and added prebiotics. That is closer to 15 percent different than 15 times different, and it produced a company that generated enough attention to land a deal on Shark Tank.
What happened after that deal is as instructive as the deal itself. Ellsworth says that after securing the Shark Tank agreement, the company took nine months off. The pause, in her telling, was deliberate: a period to get the brand and its target consumer clearly defined before returning to market. The willingness to stop momentum in order to sharpen precision is a different operating philosophy than the growth-at-all-costs posture that 10x thinking tends to produce.
The best brands and the best businesses are 10 to 15% better than the established norm. Allison Ellsworth
Matteo Franceschetti, who leads the smart mattress company Eight Sleep, gestures at a related efficiency logic in the physical product space. His projection: “250 people making a billion.” The figure is notable not for the revenue number alone but for the ratio. A billion-dollar consumer hardware business run by 250 people implies a level of operational focus that has little room for sprawl. It suggests that a refined consumer proposition, executed with discipline, can achieve scale without the headcount bloat that has become the default model for hypergrowth companies.
The mechanism Ellsworth is describing is what makes the 10-to-15-percent frame worth taking seriously across categories: it works with consumer psychology rather than betting against it. Improving a familiar product by a meaningful but not radical margin taps into existing habits rather than requiring new ones to form. That is a structural advantage, not a consolation prize for founders who lack ambition.
What is less settled is how widely the frame transfers. Ellsworth is speaking from a specific category, beverages, where familiarity and habit are especially powerful forces. Physical products with high switching costs or strong network effects may respond to different dynamics. And the claim that incremental improvement outperforms radical innovation across consumer goods broadly is a large claim that the available evidence supports but does not prove at scale.
Still, the Ellsworth argument deserves more serious treatment than it typically gets in a venture-inflected discourse that rewards ambition in its claims as well as its products. The 10x standard has produced important technologies. It has also produced a long trail of consumer products that were genuinely novel and genuinely ignored, because novelty alone does not move someone to change what they buy. The companies Ellsworth is describing, ones that find the seam between familiar and meaningfully better, are not settling for less. They are picking a harder target than it looks.