Every month, headlines cite "the government's" grocery prices — a dozen eggs at $X, a gallon of milk at $Y. Those numbers come from a real survey with a real methodology, run by the U.S. Bureau of Labor Statistics (BLS). Understanding how it works — and what it isn't designed to do — explains why CrowdCPI exists as a complement to it, not a competitor.
What the BLS retail price survey actually does
The BLS employs data collectors who visit or call a sample of retail outlets — grocery stores, supermarkets, warehouse clubs, and others — spread across a defined set of geographic sampling areas that represent the country as a whole. For each outlet, they record the price of very specifically defined items: not just "eggs," but "Grade A, large, white eggs, one dozen." That specificity is what makes the price comparable from one outlet, and one month, to the next.
Those prices are collected primarily to build the Consumer Price Index (CPI), the headline inflation number. But as a byproduct, the BLS also publishes a simpler set of numbers called the Average Price Data series, often labeled with codes starting "APU." Unlike the CPI, which is an index measuring relative change, the APU series is a literal national average price in dollars and cents for a specific item, unweighted by how much of it people actually buy. This is the number most grocery price headlines are quoting, and the one CrowdCPI's charts compare against.
Why a national monthly average is useful — and where it falls short
For its intended purpose, the BLS survey is excellent. Policymakers, the Federal Reserve, and economists need a consistent, nationally representative number to track inflation and set policy. A monthly, professionally sampled average is the right tool for that job.
It's a much less precise tool for answering an individual's actual question: "Is what I'm paying for eggs normal?" The national average blends together warehouse clubs and convenience stores, rural Idaho and Manhattan, the first week of the month and the last. Your receipt reflects one store, one ZIP code, one day. Both numbers are accurate — they're just answering different questions.
Where CrowdCPI fits in
CrowdCPI doesn't run a survey or claim to replace the BLS's methodology — we don't have the same rigor or sample design. What we do is collect real receipts and self-reported prices directly from consumers, in real time, with a ZIP code attached. That data can't tell you the "official" national average — that's the BLS's job, and we display their number right alongside ours for exactly that reason. What it can do is show the gap between the professionally sampled national figure and what real people, in real places, are being charged this week.
Think of it less as government-data-versus-crowdsourced-data and more as two instruments reading the same weather from different altitudes. The BLS number tells you the climate. Consumer submissions tell you what it feels like outside today, where you live.