costinflation/637k-prices-coffee-wars

637,374 prices: 6 categories, 12 U.S. ZIPs, 29 days

637K Prices: Coffee Wars

[image]

How do ground coffee, whole beans, pods, instant coffee, ready-to-drink coffee, and cold brew compare across U.S. ZIP markets when package quantities are standardized within each category?

This fixed research snapshot contains 637,374 unaggregated, quality-filtered price observations across 6 coffee categories, 12 U.S. ZIP markets, and 29 consecutive dates from July 21 through August 18, 2026. The analysis-ready CSV preserves product titles, dates, geography, listed package quantities and prices, plus source-neutral category-specific comparable-price fields.

What “quality-filtered” means here: the rows are unaggregated product-level observations after category, safety, and exact-duplicate filtering. The file includes derived comparison fields and is a prepared research snapshot rather than an aggregate index.

Dataset at a glance

What one row represents

One row is one quality-filtered product-title, ZIP-market, date, package-quantity, and listed-price observation in one coffee category. It is not a sale, order, demand, inventory, market-share, or consumer record. Multiple distinct offers can share a title, ZIP, and date; exact duplicate public rows have been removed.

Comparable-price method

price_amount preserves the listed price. Where a compatible package quantity is resolved, normalized_price_amount scales that price to the category-specific target shown in normalized_quantity_value and normalized_quantity_unit.

Comparison targets differ by category. Compare normalized values within the same series_id; do not compare normalized prices directly across categories with different targets or quantity dimensions. Rows whose quantity is unresolved or incompatible retain their listed price and leave normalized_price_amount blank.

Category reference

Columns

Project ideas

• —Compare how package-standardized price distributions differ among ground coffee, whole beans, pods, and instant coffee.
• —Measure geographic variation within each coffee category across the 12 ZIP markets.
• —Track daily within-category medians and dispersion through the 29-day observation window.
• —Compare ready-to-drink coffee with iced coffee and cold brew without mixing their product forms.
• —Model which package formats are most likely to yield a compatible comparison quantity.

Quick start with Pandas

import pandas as pd

df = pd.readcsv( "costinflation-coffee-wars-prices-2026-07-21-to-2026-08-18.csv", dtype={"geographyid": "string", "seriesid": "category"}, ) ground = df.query("seriesid == 'groundcoffee'").dropna( subset=["normalizedpriceamount"] ) print(ground.groupby("geographylabel")["normalizedpriceamount"].median())

Quality and interpretation

• —All 29 expected dates, all 12 ZIP markets, and all 2,088 category × ZIP × date cells are present.
• —Exact duplicate public rows: 0.
• —Formula-leading product titles: 0.
• —Every nonblank comparable price has a positive resolved quantity compatible with its category target.
• —Product titles are descriptive text, not stable public product identifiers.
• —ZIP labels describe selected markets rather than citywide, metropolitan, state, or national estimates.
• —Shipping, tax, redeemed promotions, purchases, product performance, and coverage outside the selected ZIP panel are outside this dataset.

Provenance

CostInflation prepared and released this fixed, source-neutral snapshot. Category assignment, geographic labels, exact-duplicate filtering, safety filtering, and package-equivalent calculations were applied consistently before publication. The public file contains only the 16 documented research columns.

License

Released under CC0 1.0 Universal for unrestricted reuse. Attribution is not required, but citation helps others find the release.

CostInflation Team. (2026). 637K Prices: Coffee Wars. CC0 1.0 Universal.

Released by CostInflation Team.