My mother always told me not to grocery shop hungry. Was she right?

Pia has told me forever: do not grocery shop hungry.

You will buy more than you need. You will come home with things you would never otherwise buy. Eat something first.

This belongs to the category of maternal advice so universally accepted that fact-checking it feels vaguely disrespectful.

Researchers did it anyway.

And Pia is not exactly wrong. Hunger can matter, particularly around the foods you would expect it to. But in a new study of actual grocery shoppers, hunger was not the big predictor of unplanned buying across the basket.

Whether someone had come to browse was.

And the difference was enormous.

They waited for the receipts

For a 2026 field study, researchers studied 323 people shopping at actual supermarkets, measured their hunger on the spot and individually coded 3,819 items from their receipts as planned or unplanned.

Across the full basket, hunger was not significantly associated with unplanned purchasing. Shopping purpose was.

8.6× THE ODDS
of an unplanned purchase compared with routine shoppers.

Eight point six times. That is not a modest difference hiding in a footnote.

The dangerous thing to bring into a grocery store may not be an empty stomach. It may be an empty afternoon.

Pia was not completely wrong

Before I call my mother with this information, there is an important qualification.

Hunger did have a specialty. It was associated with more unplanned purchases in snacks and confectionery, particularly among younger shoppers.

So if you enter a grocery store starving and emerge with a KitKat you had not previously identified as essential to the household, Pia still has a case.

What the study did not find was that hunger broadly explained all the other unplanned things accumulating in the cart.

Hunger, it turns out, may have a department. Browsing has the building.

Which explains a very strange $7.46

Another grocery-store experiment makes the possible mechanism easier to see.

Researchers gave shoppers mobile coupons. Some received a coupon for an unplanned category relatively close to the route they were already expected to take. Others received one that required a larger detour through the store.

$21.29
average unplanned purchases — farther route

$13.83
average unplanned purchases — nearer route

+$7.46
THE DIFFERENCE

The offer was not the surprising part. The walk was.

The coupon got them moving. The rest of the store took it from there.

Most of the store never gets a chance

That works because shoppers typically cover less than half of the areas in a grocery store.

Most products never get rejected. They remain strangers.

You cannot remember that you are nearly out of cinnamon if nothing reminds you cinnamon exists. You cannot decide the tiny sparkling waters would be perfect for guests you are not currently expecting unless the sparkling waters get an opportunity to raise the subject.

Researchers have a wonderfully joyless term for some of this: external memory cues.

Oh right. We need that.

Sometimes we really do. Sometimes the item has been a household necessity for approximately eight seconds.

Maybe the expensive thing you bring into a grocery store isn’t hunger.

It’s time.

We tend to describe impulse buying as a failure of restraint.

I saw the cookie, I wanted the cookie, and I failed a character test administered by a baked good.

But the cookie first has to appear.

Browsing puts more things in front of us. One product recalls a dinner. Another reminds us that we are nearly out of something. A third solves a problem that did not exist until we saw the solution.

Some of those discoveries are genuinely useful. That is part of what makes grocery stores enjoyable.

A list tells you what you came for. Time gives everything else a chance to make its case.

Which may explain how I grocery shop now

The irony is that I barely grocery shop this way anymore.

I order almost everything from Whole Foods for delivery. I started doing it for completely practical reasons: it is easy, it saves time, and if the produce arrives looking as though it has had a difficult week, getting a refund is remarkably painless.

I still like grocery stores. But when an order cannot get here that day, I usually go to a Whole Foods I know well enough to find the Parmesan without consulting an overhead sign.

I know where things are. I get them and leave.

I had always thought of this simply as efficiency. After reading the research, I wonder if it also changes the kind of shopper I become.

When I order from my kitchen, I am mostly replenishing. When I wander through a store, I am available for ideas.

Those are very different states.

The internet has noticed

Of course, online retailers understand what disappears when the physical aisle does: all those accidental introductions.

So they rebuild them.

Buy Again. You might also like. Frequently bought together. Sponsored. Recommended for you.

You can almost hear Amazon clearing its throat as checkout approaches. Before you go…

The aisle did not disappear. It learned your name.

There is real evidence behind that feeling. In a randomized field experiment at an online book retailer, personalized recommendations increased purchase propensity by 12.4%, largely by expanding the set of products consumers considered.

That study was not about Amazon or groceries. But the mechanism is familiar: the retailer does not need you to wander past more products if it can place more plausible candidates directly in front of you.

The physical store creates a detour. The digital store can personalize one.

But something really does change online

Researchers analyzing approximately two million online and brick-and-mortar grocery trips found that comparable routine online baskets contained 9.6% fewer categories and 14.1% fewer distinct items than in-store baskets.

Consecutive online orders were also substantially more repetitive. At the individual-item level, successive online trips contained roughly twice as many of the same items as successive trips to physical stores.

Which sounds a lot like the way I shop.

I am not strolling electronically through seafood wondering what looks good tonight. I am replenishing the things we use. The app remembers them. I remember them. Everybody is behaving very efficiently.

The physical supermarket is exceptionally good at introducing us to things. The digital supermarket may be exceptionally good at helping us repeat ourselves.

Then, just when the repetition becomes a little too efficient, the recommendation engine appears. You may also like…

There is something almost elegant about the arrangement.

First the internet removes the wandering. Then it invents a personalized version of it.

Before the choice

I think I had been giving the final decision too much credit.

We imagine influence happening after a product catches our attention—when we compare the price, inspect the package or decide whether one version looks better than another.

But both kinds of store are doing something before that.

The physical store affects what I happen to pass. The digital store predicts what I might pause over.

Before I decide what I want, someone has already helped decide what I see.

That does not mean I am being forced to buy anything. It means some products receive an introduction and others never enter the room.

And an introduction, apparently, is worth quite a lot.

So was Pia right?

About shopping hungry?

Right enough that I will not be presenting her with a formal correction over dinner.

Hunger can make the snack aisle louder. It simply does not explain the entire cart as neatly as the old rule suggests.

The finding I will remember is browsing.

The extra things in the cart may not prove hunger won. Some may simply prove I stayed long enough to meet them.

That makes me feel considerably better about ordering groceries from my kitchen.

Until Amazon reminds me that people who bought the olive oil also bought six other things I had been perfectly happy not knowing existed.

The aisle, apparently, has excellent Wi-Fi.

THE NUMBER THAT CHANGED THE STORY

8.6×

In a 2026 field study, browsing shoppers had approximately 8.6 times the odds of an unplanned purchase compared with routine shoppers.

Hunger was not significantly associated with unplanned purchasing across the overall basket. It did matter within snacks and confectionery, particularly among younger shoppers.

The broader predictor was not simply how hungry someone felt. It was what kind of shopping trip they had come to make.

THE $7.46 DETOUR

FARTHER DETOUR — $21.29 average unplanned spending

NEARER ROUTE — $13.83 average unplanned spending

DIFFERENCE — +$7.46

This does not mean every extra aisle costs $7.46. It shows that, in this experiment, shoppers sent farther through the store spent more on unplanned purchases.

What this does not prove

The research does not say hunger never affects grocery purchases, that browsing causes every unplanned purchase or that ordering online makes anyone immune to impulse buying.

And 8.6× is not a universal number to apply to every shopper or store.

The field study measured shopping purpose, not the number of minutes each person had available. “Maybe the expensive thing is time” is an interpretation informed by the browsing result and the separate travel-distance experiment—not a claim that each extra minute has a fixed cost.

The narrower—and more interesting—finding is: Hunger can affect what becomes tempting. Browsing changes how many things get the opportunity to tempt us.

Sources + Methodology

Durgun, Çetinkaya and Koç, 2026. Receipt-based behavioral data from 323 supermarket shoppers and 3,819 purchased items. Hunger showed no significant association with unplanned purchasing across the overall basket; shopping purpose was the dominant predictor, with browsing shoppers at nearly nine times the odds of an unplanned purchase compared with routine shoppers.

Hui, Inman, Huang and Suher, 2013. RFID shopping-path data, modeling and a field experiment. Typical grocery-shopping paths covered less than half the areas in a store. In the coupon experiment, shoppers sent farther from their planned path spent $21.29 on unplanned items versus $13.83 among those sent nearer their route.

Chintala, Liaukonytė and Yang, published online 2023; journal issue 2024. Approximately two million online and offline grocery trips. Online baskets had 9.6% less category variety and 14.1% less item variety, and successive online trips contained roughly twice as many overlapping items as successive in-store trips.

Li, Grahl and Hinz, 2022. In a randomized field experiment at an online book retailer, personalized recommendations increased purchase propensity by 12.4% and basket value by 1.7%, with much of the purchasing effect mediated by an expansion in consumers’ consideration sets.