The Chicken Salad Index

How it's built

This is n=1: one person's salad order, at one restaurant, watched semi-regularly since 2020. It is not a survey of restaurant prices, and it should not be read as one. The point is narrower and more honest than that — it is a record of what one specific order cost, held up next to what public data says the ingredients in it should have cost.

The modeled lines are built from BLS/FRED national average prices, not Houston-specific ones — no series exists at that granularity for most of this basket, so every ingredient line runs on the same national numbers a diner in Seattle or Tampa would see, even though the actual salad is bought in Houston. Two of the produce items, cucumber and red onion, share a single "fruits and vegetables" CPI index because the BLS stopped publishing an average price for either one — dry onions in 1997, cucumbers in 2000 — so no series covers the years this page tracks. They move in lockstep in this model, which is a simplification, not a measurement of either ingredient alone. Another example of Onion price obfuscation by our government.

The full cost-stack line adds labor, occupancy, energy, and a margin/other residual on top of ingredients, each moving with its own national proxy series (restaurant-industry wages, residential rent as a stand-in for commercial rent, the CPI energy index, and CPI all-items for everything else). The share each of those takes of the menu price — 30% ingredients, 30% labor, 8% occupancy, 4% energy, 28% margin — is an assumption drawn from typical fast-casual cost structures, not a measurement of this specific restaurant's books, which are not available to this page or its author.

Some of the observed price points come from delivery-app listings rather than in-store receipts. Delivery-platform menu prices typically run 10–20% above what the same order costs walking in, so those points should be read as an upper bound on the in-store price, not the price itself.

The menu item Sean actually orders — every day — bundles the salad with a large unsweet iced tea, so every observed "salad" price on this page already includes the tea; the modeled salad figure is built the same way, from the salad and tea recipe lines together. The standalone tea price shown in the chart and the basket table is a reference point only; it is not added on top of the salad figure, and it's a size mismatch besides — the archived tea listing is size-unstated in its earliest (2021) capture and 20 oz in every capture after that, smaller than the large tea Sean actually orders.

The observed series here is archived national list prices from a single aggregator (September 2020 through June 2024), until Sean's card history begins in August 2024; from there the record is his card statements. Sean's own recollection (~$9.85 at the start, $10.55 and $11.38 along the way, last hike around December 2023) matched the archive within 0.1% and is treated as corroboration, not record. The record itself can't pin down exactly when the last hike happened — only that the national list was $12.36 in June 2024 and Sean's store was $12.55 by August 2024, unchanged since.

The remaining archived price points — other aggregators, delivery-platform listings, and archive points after Sean's card history begins — are kept on the chart as context, not as the series of record. Neither is Sean's memory: it corroborates the record above, but where it can't be verified against a receipt or statement it doesn't count as one, so it's described in this paragraph rather than plotted.

Every observed point on the chart and every fact on this page comes from files in this site's repository; nothing is invented to fill a gap, and a gap in the data shows up as a gap on the chart.