I Weighed 31 Packages in My Own Pantry. Eleven Had Shrunk.

Sometime in August a shrinkflation segment went around again. Same shape as every other one: a hand holding two boxes of cereal, a graphic, the phrase "the price stayed the same." I have watched some version of that clip maybe a dozen times over the last few years and I always come away with the same feeling, which is that I believe it and I have no idea whether it is happening to me.

So I decided to find out. I pulled the kitchen scale out from behind the stand mixer, took everything out of the pantry, and spent about five evenings across two weeks writing down net weights.

What I actually did, because the method matters more than the number

I checked 31 packaged items. Anything shelf-stable I buy on repeat: cereal, crackers, pasta, canned goods, coffee, chips, peanut butter, the tub of oats, the bag of chocolate chips I pretend is for baking.

For each one I needed a before, and this is where most home shrinkflation audits fall apart. Memory is not evidence. I used three sources, in order of how much I trust them:

1. An older package still in the house. Best case. Two physical packages, two printed net weights, no ambiguity. This only worked for four items.
2. My own archived receipts and order histories. Grocery pickup orders list the size. I had a bit under two years of those.
3. Archived product listings. Store sites and the Wayback Machine. Weakest of the three, because a listing can be stale or wrong, so I only counted these when two independent snapshots agreed.

Anything I could not source to one of those three, I threw out of the count. That killed about nine items, including a couple I would have sworn had shrunk. I want to be plain about that: the strong feeling that something got smaller is not reliable, and mine was wrong at least twice.

Eleven of thirty-one had changed size

Eleven items were smaller than the version I had bought before. One was larger, which surprised me. The rest were unchanged.

The size cuts were not dramatic individually. That is the whole design of it. Nothing dropped by a third. The typical cut was somewhere between four and nine percent, which is small enough that the box looks the same on the shelf and the pour looks the same in the bowl.

The three tells I did not know to look for

The package redesign

Every single item that shrank had also changed its artwork. Every one. I do not think that is a coincidence and I do not think it is sinister either, exactly. If you are retooling a package anyway, that is when you change the fill. But it does mean a new look on a familiar box is worth ten seconds of your attention.

The word "new"

"New look, same great taste" appeared on three of the eleven. It is doing a lot of quiet work in that sentence. Same great taste, yes. Nobody said same great amount.

The bottom of the can

Not the pantry, but I noticed it during the same stretch: a couple of the canned items had a deeper indent in the base than the older can. Same height on the shelf, less inside. I only caught it because I had both cans on the counter at once.

What the eleven percent actually cost me

Here is the part I found genuinely useful. A six percent size cut with no price change is a six-and-a-half percent price increase, and if you have eleven of those spread across a cart it does not feel like anything. It is well under the threshold where a person notices a bill went up.

Working from the items I could verify, the size changes on those eleven added something in the range of two to three dollars a week to my grocery bill for exactly the same food. Call it a hundred and something a year. That is not a crisis. It is also not nothing, and the reason it works is that no single moment of it is large enough to argue with.

This is the same reason I stopped trusting a price by itself. The price is the number that stays still while everything under it moves. Price per ounce is the number that tells the truth, and it is the one nobody reads.

The one that got bigger

A bag of frozen fruit went up, from a size I had been buying for years to a slightly larger one, at a price that had also gone up but by less. Per ounce it got cheaper.

I have thought about this more than it deserves. I think it is a competitive story rather than a generous one. Frozen fruit is a category with a lot of entrants and a lot of private label, and going up in size is a way to make a per-ounce comparison look good on a shelf tag where everybody is being compared on per-ounce.

Which is a useful thing to notice, actually. Sizes move in the direction the pressure is coming from. In a category where shoppers are comparing per-unit prices, packages grow. In a category where shoppers are anchored on a price point, the $3.99 box or the two-dollar bag, packages shrink, because the price is the thing that cannot move.

If you want to guess which of your groceries will shrink next, that is the test I would use. Ask whether you know the price of the thing or the size of the thing. Whichever one you know is the one that will hold.

What is actually driving it

I want to be fair here, because "shrinkflation" is usually reported with a tone of pure villainy and the reality is more boring.

A manufacturer facing higher input costs has three options. Raise the price, cut the size, or accept a thinner margin. The third one is not a real option for very long. So it is a choice between two ways of charging you more, and the research on this is fairly consistent: people notice a price change far more readily than a size change. Given a choice between the change customers punish and the change they do not notice, companies pick the second one, every time, and I do not think any individual is being especially wicked in the process.

What I object to is not the size cut. It is the redesign that accompanies it, which is doing something more deliberate. It removes your ability to compare the new package to the one in your memory. That is not cost management. That is specifically making a comparison harder, and it is the part that annoys me.

How to do this yourself in an evening

You do not need thirty-one items. Pick the eight or ten things you buy every single week, because those are where a size cut compounds.

Find your before. This is the only hard part. In order of trustworthiness: an older package still in the house, then your own receipts or online order history, then two independent archived listings that agree. If you cannot source it, skip the item. I know how tempting it is to count the one you are sure about. Do not.

Write down the net weight, not the price. The price you can look up any time. The size is the thing that disappears silently.

Photograph the nutrition panel. Four seconds, and it gives you a dated record instead of a feeling. This is the single habit from the whole exercise that I have actually kept.

Do the per-ounce math once, not every week. The point is not to become a person who calculates in aisles. The point is to find out which of your regulars moved, deal with those, and then go back to not thinking about it.

Put a reminder in for a year out. Mine is in September. Twenty minutes a year is a completely reasonable price for knowing rather than suspecting.

What I changed afterward

Not much, honestly, and I want to be careful not to oversell a lifestyle here. I did three things.

I started buying store brand on four of the eleven, because a size cut is a good moment to test whether you were loyal to the product or to the habit. Two of the four stuck. I wrote about how those swaps went later that fall.

I started photographing the nutrition panel of anything I buy regularly, which sounds obsessive and takes four seconds and gives you an actual dated record instead of a feeling.

And I stopped assuming the biggest package wins. It usually does. On two of the eleven, after the size cut, it stopped being true, and the mid-size package quietly became the better buy. Nothing on the shelf tells you when that flips. That is the whole reason the tool on this site exists.

Would I do it again

Yes, and I have it on the calendar for next September. Five evenings is not much for a number you can actually rely on, and I would rather have thirty-one verified weights from my own kitchen than a strong opinion I picked up from a two-minute segment.

If you do it yourself, the only rule that really matters is the one about evidence. Old package, old receipt, or two independent archived listings. If you cannot source the before, do not count the item. It is tempting, it will make your number bigger, and it will make the number worthless.

The thing I did not expect to get out of this was how much calmer I feel about my grocery bill now. Not because it went down. It did, by a couple of dollars a week, which is not life-altering. Because the vague sense that I was being got at, somehow, in a way I could not point to, turned into eleven specific items and a number. Eleven items is a manageable problem. A vague sense of being got at is not a problem at all, it is just a mood, and you cannot do anything about a mood.