Every brand competes for the same moment: the instant someone decides what to buy. For most of the last century that moment happened in front of a shelf. Today, more and more of it happens in front of a screen, in a list of results ranked by a marketplace. The shelf and the search result look like they do the same job. They reward different things, and the brands that understand the difference are the ones that grow.
The century of the shelf
The shelf is younger than it seems. Until the early twentieth century, most grocery customers did not choose products at all. They handed a list to a clerk, who walked the store and collected the goods for them. In 1916 Clarence Saunders opened the first Piggly Wiggly in Memphis, which became the first self-service grocery chain: customers walked the aisles and gathered their goods themselves. Saunders understood something that still drives retail today. When customers see more, they buy more.
That change gave packaging a new job. Once the clerk was gone, the package had to do the selling on its own, in silence, from a shelf crowded with alternatives. Names, colours and logos became the way a product introduced itself.
It also made the shelf the most contested space in retail. Positions were negotiated with retailers, and the best of them went to whoever could pay for them. For decades, a brand’s visibility depended less on how good it was than on how large it was.
The shelf moves online
Marketplaces changed who gets onto the shelf. In 1999, third-party sellers accounted for 3% of the goods sold on Amazon. By 2018 they accounted for 58%. Independent brands that could never have negotiated a place in a national retail chain now sell alongside the largest companies in the world, on the same page, in the same format.
The online shelf has no length limit, but attention still does. Shoppers look at the first results and rarely much further, so position still decides almost everything. What changed is how position is won. Advertising can still buy a place near the top, as it always could. The position that lasts, however, is earned continuously from the product’s own record: how often it sells, how often a visit becomes a purchase, what customers say about it, and whether it is in stock when they want it.
That makes the results page a system that feeds itself. A product that sells is shown more often, and a product shown more often sells more. The same logic works in reverse, which is why a stockout costs far more than the sales missed while it lasts. The position has to be won back afterwards.
What customers now read instead of a clerk
Reviews have become the clerk’s recommendation, written down and visible to everyone. Research from Northwestern University’s Spiegel Research Center found that a product with five reviews is 270% more likely to be purchased than one with none, and that the benefit of each additional review falls quickly after the first five.
The first customers of a brand therefore carry unusual weight. They are not just early revenue. Their words become part of the product page that every later customer reads.
Everything else on that page is doing the work the package once did alone. The main image competes as a thumbnail in a grid of alternatives. The title has to match the words customers type, not the words the company uses internally. The images that follow answer, in order, the questions that stop a purchase. None of this rewards size. It rewards precision and consistency, which is why a small brand can outrank a far larger one on the page that matters most.
“None of this rewards size. It rewards precision and consistency.”
The clerk returns
The next change is already under way. In 2024 Amazon introduced an AI shopping assistant inside its app, and assistants that answer shopping questions and recommend products in conversation are spreading across marketplaces and search engines.
A growing share of buying decisions may soon pass through an intermediary who chooses on the customer’s behalf, much as the clerk did before self-service. The clerk is coming back, this time as software.
An assistant cannot be charmed by a package. It recommends what it can describe with confidence: products with clear and accurate information, consistent reviews, reliable availability and an identity that makes sense. The brands being built to be legible to that kind of reader are preparing for the next shelf. The ones still optimising for the last one will find it has moved again.
How we work
I study results pages the way earlier generations of merchants studied shop windows: what catches the eye at thumbnail size, which words customers actually use, how a page answers a doubt before it becomes one. SVW’s brands are built for the shelf as it exists now and for the one that is coming, and they are operated every week against the rules that decide who is seen.
We apply the same discipline to the small number of external brands we work alongside each year. Where products are found keeps changing. The brands that earn trust, and make it easy to see, keep being found.
