You walked into a store. You picked up eggs. You looked at the price. You complained quietly. You bought them anyway.

A clean, elegant system.

Then technology arrived and asked the most Silicon Valley question imaginable: "What if the price tag knew things about you?"

In December 2025, Consumer Reports and the Groundwork Collaborative ran 437 shoppers through Instacart in four cities, all adding the same items to their carts from the same stores. Nearly three-quarters of the products came back at more than one price. A dozen eggs at a D.C. Safeway showed up at five different prices — same store, same time, different people.

That is not inflation. That is the cash register doing personality analysis.

Two ideas worth separating here.

Dynamic pricing says: "The price changed because demand changed." Surveillance pricing says: "The price changed because you showed up."

Airlines have done dynamic pricing for years. Annoying, but the explanation is usually demand, timing, and whether you made the terrible mistake of needing to fly during March Break.

Surveillance pricing is different. That's when companies use data about you — browsing history, device, location, loyalty profile, purchase history, urgency — to estimate what you personally might tolerate paying.

Not "what is this worth?" but "what can we get from Ross?"

Lovely.

Now, credit where it's due: Instacart pulled the tests and insists they never touched personal or behavioural data. I'll take them at their word. But notice how thin that is. The lab closed after someone pulled back the curtain — not before. The infrastructure that varied a carton of eggs five ways still exists. "We didn't use your data this time" is not a guarantee. It's a status update.

Walmart got pulled into the same conversation through its rollout of digital shelf labels, on track for every U.S. store by year's end. Walmart says the tags improve accuracy, keep prices identical for everyone in a store, and collect no shopper data. Fine. Nobody is against employees spending less time changing tiny paper tags like medieval monks updating cereal prices by candlelight.

But digital labels make price changes fast, centralized, and scalable. And once prices can change instantly, the obvious question is: based on what? Demand? Time of day? Location? Loyalty data? Browsing history? Whether you searched "baby formula near me" at 2:14 a.m. like a hostage negotiator?

That is exactly why a U.S. senator wrote to Walmart's CEO and asked, in writing, whether browsing history, purchase history, device, or geo-location is used to set prices. There's now a bill that would ban electronic shelf labels in any grocery store over 10,000 square feet and prohibit surveillance pricing outright.

When Congress starts asking whether your grocery store is pricing bananas off your metadata, we've wandered slightly off the innovation brochure.

Retailers will call this personalization. Of course they will. "Personalization" is what companies say when "we built a machine to find your breaking point" tests poorly with focus groups.

The problem isn't that prices move. Prices have always moved. The problem is that prices are becoming private.

A public price tag says: "This is what the product costs." An algorithmic price says: "This is what we think you'll pay."

That's a massive shift in power. You cannot inspect the model. You don't know what data was used, what category you were placed in, or whether the price reflects supply, postal code, device type, or the fact that you've bought the same coffee every week for four years and the algorithm knows you're emotionally dependent on it.

Which, to be fair, it's right. But still rude.

And this isn't a grocery issue. It's the next stage of the data economy. First platforms collected your data to target ads. Then to recommend content. Then to shape behaviour. Now to price reality.

Here's the uncomfortable part for businesses and governments: if consumer platforms can do this with groceries, what happens when the same logic moves into enterprise software, cloud, procurement, insurance, and AI? What happens when vendors know your renewal pressure? Your dependency on one platform? Your migration difficulty? Your inability to leave?

That is not customer insight. That is leverage.

This is why data sovereignty is not some policy slogan wrapped in a flag and left to nap in a government presentation. Data sovereignty is control — over who can collect your data, combine it, infer from it, and use it against you. Because once enough behavioural data is assembled, the question is no longer whether your information is "secure." It's whether it has become a pricing weapon.

At SkyeConnex, our position is simple: no single provider should hold the whole picture. Not the whole file. Not the whole key. Not the whole context. Not the whole behavioural map.

Because the future of AI, commerce, and cloud won't just be about where data is stored. It'll be about who can assemble enough of it to predict you, price you, and push you.

Dynamic pricing used to mean flights got expensive at Christmas. Now it means a senator has to write a letter to ask whether your browser history helps decide your grocery bill.

Progress, apparently.

The shelf tag got smarter. The customer did not get more powerful.

That's the problem.


Sources

Consumer Reports / Groundwork Collaborative — Instacart pricing investigation (Dec 2025): https://github.com/consumerreports/instacart

CNBC — study on Instacart AI pricing: https://www.cnbc.com/2025/12/09/study-instacart-ai-pricing-cost-of-groceries.html

CBS News — Instacart ends AI price tests: https://www.cbsnews.com/amp/colorado/news/instacart-ends-ai-price-testing-tool-eversight

U.S. Senate letter to Walmart on dynamic pricing and customer data: https://www.banking.senate.gov/imo/media/doc/walmart_dynamic_pricing_letter.pdf

CNBC — Walmart digital shelf labels in every U.S. store by end of 2026: https://www.cnbc.com/2026/03/21/walmart-digital-price-tags-will-be-in-every-us-store-by-end-of-2026.html

Bankrate — Stop Price Gouging in Grocery Stores Act and surveillance pricing: https://www.bankrate.com/personal-finance/digital-price-tags-efficiency-or-price-gouging/


Originally published by Ross Norrie, founder of SkyeConnex, on LinkedIn.

Published May 28, 2026 · More from the SkyeConnex blog