Our Methodology
Shrinkflation is hard to prove from memory alone. This page explains exactly how ShrinkWatch gathers its data, how we verify it, and how every number on the site is calculated โ so you can judge the evidence for yourself.
Where the data comes from
Every entry on ShrinkWatch is a dated size report: a record that a specific product changed from one package size to another, in a given year, ideally with the price at the time. Reports come from three places, and we treat them differently:
- Community submissions. Shoppers submit changes they have personally observed, along with a source where possible (a photo, an old receipt, an archived product page, or a retailer listing). These are the backbone of the site.
- Public records and archives. Where a submission can be checked against a manufacturer's own past packaging, a retailer's cached listing, or a widely-reported change, we cite that alongside the report.
- Our own findings. Some entries come from directly comparing current packaging against documented earlier sizes.
We do not scrape prices in bulk or estimate sizes with a model. A product only earns a page once there is at least one concrete, dated report behind it.
How we verify a report
A claim that "this got smaller" is only useful if it can be checked. Before a report is treated as documented, we look for:
- A specific old size and new size in the same unit (ounces, grams, count, sheets), not a vague "it seems smaller."
- A date or year for the change, so the history is chronological rather than anecdotal.
- A source or corroboration where one exists โ a receipt, a photo, an archived listing, or multiple independent reports of the same change.
Reports that can't be substantiated, or that turn out to be a normal product-line change rather than a quiet downsizing, are corrected or hidden. If you spot something wrong, the report-a-problem page goes straight to us.
The math behind the numbers
Two figures do most of the work on ShrinkWatch. Both are deliberately simple so you can reproduce them yourself.
Shrink percentage
(old size โ new size) รท old size ร 100. A cereal box going from 18ย oz to 15.5ย oz has shrunk 13.9%. When a product has several reports over time, the headline figure is the cumulative change from the earliest documented size to the most recent one โ not just the latest single step.
Price per ounce
price รท new size in ounces. This is the number that exposes stealth price increases: when the size drops but the price holds, the price-per-ounce climbs even though the shelf tag looks unchanged. We only show it when a price was actually recorded with the report โ we never invent one.
You can run these calculations on any product yourself with our price-per-ounce calculator.
What the data can and can't tell you
We would rather be honest about the limits than overstate them:
- Sizes and prices vary by store, region, and promotion. A report is a real observation at a point in time, not a universal price.
- A documented size drop is not automatically wrongdoing. Recipes, regulations, and product lines change for legitimate reasons too. We record what changed; you decide what to make of it.
- Coverage is uneven. Products with only a single report show that one data point and are marked as such โ they are a starting record, not a full history.
You can see the current size of the dataset any time on the stats page.
Help make it better
ShrinkWatch gets more accurate every time a shopper adds a dated, sourced report. If you have noticed a package quietly shrink โ or kept an old one you can measure โ you can submit a size report in a couple of minutes. The more corroboration a change has, the more confidently it holds brands accountable.
See the data for yourself
Browse documented size changes across brands and categories.
Browse tracked products โ