Piggly Wiggly triggered the explosion of consumer marketing. What will the demise of Google search give us?

In 1916, Clarence Saunders opened a grocery store in Memphis called Piggly Wiggly. It's widely regarded as the first self-service grocery store.

Before this, the shopkeeper was how products got sold. Suppliers didn't advertise to consumers — they tried to influence the people behind the counter, because they were the ones who recommended what to buy. Wholesale relationships, trade discounts, the right word across the counter. That was how products moved.

Piggly Wiggly removed the person behind the counter.

Once they were gone, manufacturers had a problem: how do you sell something when nobody's recommending it?

The answer, it turned out, was packaging. And branding. And advertising. And the entire consumer marketing industry as we know it today.

Without the shopkeeper's recommendation, the package itself had to do the selling — visually, immediately, in the three seconds a shopper walked past. Goods that had come in unmarked bulk containers now had to arrive individually wrapped, clearly labelled, visually distinct. If your product didn't catch the eye, didn't communicate its benefit at a glance, it didn't sell.

Kellogg's, Heinz and Campbell's had all existed for decades before Piggly Wiggly opened. What changed was how they had to compete. The brands that understood this and invested in packaging, advertising and direct consumer appeal pulled away from those that didn't. Campbell's launched one of the first mass advertising campaigns in this period; streetcar ads doubled their city sales. Kellogg's overhauled its packaging and processing. The brands that had relied on trade relationships and the shopkeeper's word found those advantages gone overnight.

The wholesale distributors who'd built their margins on being the conduit between manufacturer and shopkeeper largely disappeared in the decade that followed. Not because Piggly Wiggly wiped them out directly. Because the model they served no longer existed.

How fast did Piggly Wiggly grow? Within four years — by 1920 — it had hundreds of stores and revenues of $60 million a year. In today's money, roughly $970 million. By 1922, 1,200 stores across 29 states.

For a sense of scale: Lovable — the AI-native software builder I wrote about in Post 2 — went from launch to $100 million ARR in eight months. Both stories share the same quality: from nothing to enormous, very fast, built on technology that already existed. Saunders didn't invent the turnstile or the price label or the shopping basket. He put them together in a way that removed the human in the middle — and the market responded violently.

(Saunders then lost the company in a stock market speculation gone wrong in 1922. Piggly Wiggly settled into being a regional grocery brand. Lovable may yet have its own version of that story. Fast growth and structural challenge have always arrived together.)

Now here's why I think this matters in 2026 — and it's a closer parallel than it might first appear.

For the past 25 years, Google has been the shopkeeper. Not a person, but the same function: standing between a brand and the customer, deciding what to recommend. Businesses didn't need to reach customers directly — they needed Google on their side. An entire industry — SEO, content marketing, Google Ads — was built purely to win that recommendation.

The shift toward AI-powered search — across ChatGPT, Perplexity and Google's own AI features — has fundamentally changed who gets the click, and who doesn't.

The most striking example is HubSpot. HubSpot is a software business — CRM and marketing tools — that built its customer acquisition around one of the most sophisticated content marketing machines in the world. Thousands of articles, guides and tools, all optimised to rank on Google and funnel readers toward their products. At peak: 24 million visits a month to their own blog. By early 2025, around 6–7 million. Their own CEO said it on an earnings call: "AI overviews are giving answers, and fewer people are clicking through to websites."

If it can happen to HubSpot — with one of the best SEO teams on the planet — it can happen to anyone who built their discovery model around winning a recommendation from an intermediary.

The industry's response has been fast. AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization) appeared almost overnight. Adobe, Semrush, HubSpot itself — all launched or acquired products in this space. Dozens of new agencies appeared within months.

But I can't help thinking that trying to influence a different shopkeeper may not be the solution.

The problem is the same as it was in 1916. Optimising to win an AI's recommendation is still a recommendation-dependent model. You're building visibility around an intermediary you don't control — one whose preferences you can only guess at, whose algorithm can change overnight, and who has no particular reason to keep you in the answer.

So what happens next?

In 1916, the answer was clear: the consumer is now doing the choosing, so market directly to the consumer. Build packaging they'll reach for. Advertise to people, not to shopkeepers. The entire modern brand industry grew from that single insight.

But in 2026, it's not obvious that the consumer is the one doing the choosing. What if — as some are already predicting — your AI starts talking to my AI and just buys things? What if the agent doing the choosing isn't a person at all, but a system briefed by a person, acting on their behalf?

At that point the 1916 parallel forks.

One path says: the consumer is still in charge — they're just using an AI to act for them. So you still need to win human preference. Build brands people want and instruct their agents to buy. The answer is still consumer marketing, just upstream of the transaction.

The other path says: if the agent is making the final call, you need to make the agent want you. Not the human — the machine. Which means something that doesn't really exist yet. AI brand marketing, built not for human eyes but for algorithmic preference. Nobody has written that playbook.

The AEO and GEO industry thinks it's on path two — optimise for the AI recommendation. But it's still building for a world where an intermediary recommends and a human decides. What happens when the intermediary both recommends and decides?

Kellogg's, Campbell's and Heinz grew into household names by building brands that worked without a shopkeeper. The brands that thrive in the next decade may need to build something that works without a human in the decision loop at all.

Nobody had a playbook for the post-shopkeeper world in 1916. They built one.

Nobody has one for this yet either.

Sources: Tennessee Encyclopedia (Piggly Wiggly); Encyclopedia.com / FundingUniverse (store count and revenue data); BLS CPI data via in2013dollars.com ($60M 1920 = ~$970M 2025); SparkToro/Semrush (zero-click search data, 2025); HubSpot Q4 2024 earnings call (CEO Yamini Rangan); Ahrefs/Semrush (HubSpot traffic data, Nov 2024–Jan 2025); American Business History Center (Campbell's advertising history); TechCrunch (Lovable, Dec 2025); Adobe/Semrush/Scrunch (AEO/GEO industry, 2025).

Article 7 of the "Every Era Has Its Reboot" series — exploring what every major technology shift actually meant for ordinary businesses, and what AI means for yours. | redebuter.com