Part 6 of The $200 Grocery Bill: An Uncomfortable Look at Why American Food Costs So Much

Investigative satire. Skeptical of everyone. Funny, but fact-based. Nobody gets off easy — including you.

Part 5 showed you the toll booths between the farm and your cart. Today: where the money actually went — and why grocery store advertising is now growing ten times faster than the grocery store.

Here’s the most useful sentence in this entire series. It isn’t mine. It isn’t complicated. It fits on a bar napkin.

You can only spend a dollar once.

That’s it.

That’s the whole thing. Six articles, forty thousand words, a hundred and fifty citations, and every one of them is a footnote to a sentence a nine-year-old could check.

Every dollar in American grocery comes from exactly one place. A person. Standing at a register. On a Tuesday.

And then it gets spent. One time. On one thing.

It can go into the price on the shelf tag.

Or it can go into grocery store advertising. Or slotting fees. Or a display allowance. Or a chargeback for the wrong door. Or a media buy. Or a data platform. Or — and I want you to hold onto this one, because we’re coming back to it — a television manufacturer.

Not both.

Never both.

And in thirty years in this industry I have never once heard anyone in a hearing room, a boardroom, or a press release be asked to say which one they picked.

The Industry Has an Answer, and It’s Nonsense

Let’s do the counterargument first, because it’s the one you’ll hear the moment you raise any of this.

McKinsey’s finding, from a 2022 piece titled — and I promise I am not making this up — “Busted! Five myths about retail media”: “more than 80 percent of spend flowing into retail media networks is incremental and comes from all sources.”

Incremental. Fresh money. Nobody’s discount got taken away, nobody’s price went up, everybody wins.

If that’s true, this article is over. Go read something else.

So look at what that actually is.

It’s a survey.

Of the people spending the money.

Asking them whether the money they spent was money they’d have spent anyway.

It cites no sample size. It names no methodology. Its evidence is the phrase “our research suggests.” And it is four years old, which in this business is a geological era.

That’s the myth-busting.

I’ve spent thirty years in food and beverage. I have sat in the rooms where those budgets actually get built, on both sides of the table, with the coffee and the deck and the guy from finance.

Here is my professional assessment of that finding:

It’s bullshit. You can only split one dollar a hundred ways.

Every consumer packaged goods company on earth has exactly one pile of money.

One.

Whatever went into retail media came out of something. Trade promotion. Brand marketing. Headcount. Packaging. R&D. Margin.

That’s the list. That is the entire list. I’ll wait while somebody adds to it.

If it came out of trade promotion, it came out of your discount.

If it came out of margin, it went into the case cost — and Part 5 showed you what happens to the case cost. Marked up about thirty percent. Handed to you at register six.

They Have Never Sat In Your Seat

And here’s the part I want to say carefully, because it isn’t an insult and I don’t mean it as one.

The people who write these studies are smart. Often brilliant. Disciplined in a way most operators aren’t, and better at structuring a problem than anybody I’ve worked with.

They have also, almost never, been operators.

They come out of the best programs in the country, into a firm, into a conference room. What they produce is a mathematical formulation. And on paper, it is right. The logic closes. The arrows point the right way. The deck is beautiful.

Then it meets Tuesday.

Tuesday is a truck that didn’t show. Tuesday is a buyer who got promoted and his replacement doesn’t like your category. Tuesday is a co-packer with a line down and a promotion that ships in nine days. Tuesday is the thing no model has a cell for.

I have sat across from major-league consultants as the operator — the guy who had to go make the plan work — and been told things that could not have been further from the truth. Not because they were lying. Not because they were stupid.

Because they had never sat in my seat.

And there’s a difference between a person who has modeled a decision and a person who has had to live inside one for four quarters.

(Full disclosure, since this series doesn’t let anybody off, including me: I run a consultancy. I’m one of them now. The only distinction I’d claim is thirty years of having been the guy on the other side of the table first, and a standing rule that I don’t tell an operator anything I haven’t had to do myself.)

There is a famous illustration of this, and it’s McKinsey’s own.

In 1982 two McKinsey consultants published In Search of Excellence, profiling the great American companies and what made them great. It became one of the best-selling business books ever written and effectively built the modern consulting industry’s reputation.

Two years later, BusinessWeek ran a cover story on how those excellent companies were doing.

The headline was “Oops! Who’s Excellent Now?”

Two years. That’s how long the model held.

So when a consultancy tells me that eighty percent of retail media money is brand new — money that appeared, that displaced nothing, that came out of no other line — I don’t think they’re lying to me.

I think they’ve never had to build the budget it came out of.

There is no third door.

There is no line on any income statement in the United States of America that reads “new money.” There is no vault. Nobody found a dollar in the parking lot.

A brand can call it whatever it likes on the budget spreadsheet.

The dollar does not care what you named the column.

Where the Dollar Went Instead

So where did it go? The filings are not shy.

Start with the biggest grocery store advertising operation in the country.

Walmart’s global advertising business grew 46 percent last fiscal year, to nearly $6.4 billion.

Walmart’s total revenue that year grew 4.7 percent.

Read those two numbers again. The advertising operation inside the store is growing about ten times faster than the store it’s inside.

That’s not a business unit. That’s a tenant eating the building.

And grocery store advertising is not a rounding error anymore — the category ran about $58.8 billion in 2025 and is forecast at $69.3 billion this year.

And in December 2024, Walmart bought Vizio — the television manufacturer — for about $2.3 billion. Not to sell televisions. It already sold televisions, at a famously miserable margin. It bought the operating system inside them. Walmart’s EVP and Chief Growth Officer for Walmart U.S., Seth Dallaire, said in the closing announcement that Vizio had “expertly changed their business over time, like building and quickly scaling a profitable advertising business.“

The largest grocer in America bought a television company and announced, in the press release, that the appealing part was the advertising.

I’m not inferring that. I’m not reading between lines.

That’s the quote.

Kroger’s “alternative profit” businesses — media, data, personal finance — delivered $1.5 billion in operating profit in fiscal 2025. Kroger’s total operating profit that year was $1.9 billion.

Look at those two numbers.

They’re nearly the same number.

Only one of them involves groceries.

Now the honest asterisk, because I’d rather hand it to you than have a fact-checker hand it to you. That $1.9 billion is beaten half to death by one-time charges — chiefly a $2.5 billion impairment on Kroger’s automated warehouses. On adjusted operating profit of $4.9 billion, the data-and-media side is closer to thirty percent than eighty.

Thirty percent.

I’ll take thirty percent. Thirty percent of a company’s operating profit, from a business that sells no food, at a company whose entire public identity is selling food.

Nobody calls that a side hustle either.

And McKinsey — the same people with the survey — describe retail media as having shifted “from being about selling ads to becoming part of the grocery profit architecture.“

Architecture.

Not a pilot. Not a side project. Not an experiment somebody’s running in three markets.

Load-bearing.

Which is a hell of a word to pick, when you stop and think about what the building was originally for.

Three Words Part 4 Missed

I took the vocabulary apart in Part 4 — “loyalty,” “rewards,” “guest,” “retail media network,” “shopper insights,” “alternative profit businesses.” I’m not running those again.

But I missed three, and they’re the three that matter for today, because all three exist specifically to avoid naming a dollar.

“Shopper marketing.” Marketing to the shopper, obviously. That’s how it’s meant. Read it as a possessive, though, and it turns honest. The shopper being marketed. The shopper as the raw material rather than the audience.

“Commerce media.” Pure vapor. Two abstractions holding hands so neither one has to say what happened. Nothing in that phrase indicates that a human being bought a jar of anything, which is presumably the point.

“Trade spend.” And here’s the masterpiece. Hundreds of billions of dollars a year, and the word we use for it is “spend” — a verb, pressed into service as a noun, so that nobody ever has to say what it was spent on, or what it was spent instead of.

It’s a word with the object surgically removed.

Now go back through the whole glossary, Part 4’s and mine together. Loyalty. Rewards. Guest. Retail media network. Shopper insights. Alternative profit. Shopper marketing. Commerce media. Trade spend.

Find me the word for the shopper’s discount.

There isn’t one.

There is no term of art in this entire industry for the price that didn’t come down. It’s the only thing in the system nobody bothered to name — because it’s the only thing nobody ever has to report.

What It Would Have Been Worth

Let’s put a number on it, carefully, because “the money could have gone to the shelf tag” is a sentence and not a figure.

US retail media advertising ran about $58.8 billion in 2025, and is forecast at $69.3 billion for 2026. Amazon alone books $68.6 billion in advertising services. And of the roughly ten and a half billion dollars in new retail media money expected this year, Amazon and Walmart together are projected to take more than 89 percent of it — which, if you read Part 3, is the same shape as everything else in this industry: a wide funnel and a very narrow gate.

Two Million Dollars To Be Allowed To Try

Now the older money, which is bigger and which nobody counts. Grocery slotting fees, by the FTC’s own survey, run $2,313 to $21,768 per item, per retailer, per metro area — and suppliers reporting that a national new-product launch required one and a half to two million dollars in slotting before a single jar sold.

Two million dollars to be allowed to try.

I can’t give you the industry total, and neither can anyone else, because — as Part 5 established — not one major American grocer discloses a dollar of it. The largest single input into what you pay is unmeasurable by design.

But you don’t need the total to see the shape. On our unit of measure, the $200 cart, the store keeps about four dollars and twenty cents. The advertising business inside that same store is growing ten times faster than the store. And every dollar of it is a dollar that arrived at a register and then went somewhere that wasn’t the shelf tag.

A Deposition Regarding the Dollar

Lenny Bruce used to read his own court transcripts on stage, deadpan, because the transcript was funnier than anything he could write. In that spirit, I’d like to depose a dollar bill. The witness has been sworn. The witness is being very cooperative.

Q: State your denomination for the record.

A: One dollar.

Q: Where were you last Tuesday, around four fifteen in the afternoon?

A: In a woman’s hand. Register six.

Q: Do you know her name?

A: We weren’t introduced. She was buying pasta sauce.

Q: And what have you been doing since?

A: Oh, busy week.

Q: Busy.

A: Very productive, actually.

Q: Walk us through it.

A: Sure. I sat with the retailer a couple of days. Then I went out to the sauce company.

Q: Why?

A: Because the retailer had bought the sauce from them. I was part of what it cost.

Q: So you left the store.

A: Briefly.

Q: Briefly?

A: I was back by Thursday.

Q: You went to the manufacturer, and then you came back to the retailer.

A: Slotting allowance. And some display money. There was also a chargeback for a late delivery but that wasn’t me, that was a colleague.

Q: Let me make sure the record is clear. The store paid the sauce company for the sauce. And then the sauce company paid the store for permission to sell the sauce.

A: Correct.

Q: Isn’t that the same money going in a circle?

A: It’s two different line items.

Q: That isn’t what I asked.

A: It’s two different departments.

Q: Where are you now?

A: Media, I think. Somebody bought the top of a search result.

Q: For the pasta sauce.

A: For the pasta sauce.

Q: She had already bought the pasta sauce.

A: Different quarter.

Q: At any point in this itinerary — the store, the manufacturer, back to the store, the endcap, the search result — did you ever become a lower price?

A: No.

Q: Was it raised as an option?

A: I don’t believe it came up.

Q: Who would have raised it?

A: …

Q: Take your time.

A: That’s sort of the thing.

Q: Say it.

A: There’s a Chief Growth Officer. There’s a Chief Merchandising Officer. There’s a VP of Shopper Marketing, a director of category management, a trade promotion analyst, a media buyer, and about forty people at the agency.

Q: And?

A: In the whole trip, I never once passed anybody whose job it was to be her.

Q: Nobody represents the woman at register six.

A: There isn’t a role for it. I looked.

*Q: One final question. Could you have done all of that and been a discount?*

A: No.

Q: Why not?

A: Because I’m a dollar. I’m not a metaphor. You spend me, I’m gone.

Q: Thank you. Nothing further.

Let the record show the witness may be spent.

The Verdict

Stack it up.

A television manufacturer, bought for two point three billion dollars, so that grocery store advertising could follow you into your living room. An ad business growing ten times faster than the store it lives in. A data subsidiary. A word — alternative profit — that concedes in an SEC filing which business is now the main one. A vocabulary with nine terms for taking money and not one for the discount that didn’t happen.

Every one of those is a dollar.

Every one of those dollars came from a person at a register.

And not one came back as a lower price, because it had already been spent.

Now, this is the part where I’m supposed to say greed. That’s the move. It’s what you’re expecting and it would feel good, and I’m not going to do it, because it isn’t true and it lets everybody off.

Nobody had to be greedy for this.

Every one of those decisions was made by a competent adult doing their job, in a meeting, with a spreadsheet. Ask any of them and they’d walk you through the logic and you’d nod.

It’s just that money doesn’t multiply.

It goes one place. It goes there once. And when an entire industry quietly agrees — without ever holding a meeting about it — to send that dollar anywhere but the shelf tag, and when nobody in any hearing room in this country has ever made them say where it went,

you get what we’ve got.

A nation of people who can’t understand why the cart costs two hundred dollars.

Standing in front of a shelf that’s been financed, optimized, auctioned, brokered and monetized six ways, right up to the second they reach for the jar.

And the truly maddening part is that none of it was hidden.

The six point four billion. The two point three billion for a TV company. The one point five out of one point nine. It’s all filed, audited, stamped, and sitting on a government website, free, searchable by any staffer, any reporter, any regulator, any citizen with a browser and one afternoon.

Nobody hid anything.

Nobody had to.

Because the only question that would break this open has never once been asked out loud, and it is four words long, and a nine-year-old could ask it:

where did the dollar go?

Case open. Court adjourned, not dismissed.

Wednesday — Part 7: the biggest pile of money in this entire series, and the one word that gives the whole game away. Consumer packaged goods companies put about a fifth of their revenue into something called “trade promotion.” Some of it becomes your discount. Some of it becomes a warehouse full of product bought cheap and sold to you at full price five months later. There’s a technical term for whether the money reaches you. The fact that the term exists is the story.

Sources & Receipts

• Walmart FY2026 Q4 earnings release — global advertising +46% to nearly $6.4 billion; total revenue +4.7%; advertising recorded “in either net sales or as a reduction to cost of sales”
• Walmart completes acquisition of VIZIO, December 3, 2024 — ~$2.3 billion; Seth Dallaire on VIZIO’s “profitable advertising business”
• Kroger Q4 and Full Year 2025 results — $1.5 billion operating profit from alternative profit businesses against $1.9 billion total
• Amazon Q4 2025 results — advertising services net sales $68.6 billion in 2025
• EMARKETER — US retail media ad spending $58.79B (2025), $69.33B (2026); Amazon and Walmart to take $9.42B of $10.53B in new spending
• McKinsey — *The State of Grocery North America 2026*: retail media as “part of the grocery profit architecture”
• McKinsey (2022), “Busted! Five myths about retail media” — “more than 80 percent of spend flowing into retail media networks is incremental”
• FMI — the average US supermarket carries 33,248 items

Author’s Note On Method

The In Search of Excellence account — two McKinsey consultants, 1982, and BusinessWeek’s “Oops! Who’s Excellent Now?” cover two years later — is from memory of a well-known episode and should be spot-checked against the original issue before this runs.

Everything I say about what consultants get wrong is my own experience as the operator on the other side of the table. It’s testimony, not data, and I’d rather label it than dress it up.

One number in here is weaker than it looks and I’ve said so in the text: Kroger’s $1.5 billion against $1.9 billion is a comparison against a year beaten down by a $2.5 billion impairment. On adjusted operating profit the data business is nearer thirty percent than eighty. Thirty percent is still the story.

Where the evidence cuts against the argument, it’s in the text and not the footnotes. If research can’t verify it, it doesn’t ship.