Building a CPI Price Proxy from Odyssey Lite Daily Brand Sales

Official CPI arrives monthly and describes a month that has already ended. Researchers who want an earlier read on consumer prices often turn to point-of-sale data. Odyssey Lite is a simple way to start. It aggregates granular transaction rows from the Odyssey dataset into daily sales per consumer brand and company.

Each daily record carries the transaction date, brand, mapped manufacturer ISIN, unique stores, unique customers, units sold and dollar net sales. History runs from 2013-01-01 to present day, and a live feed continues from there. That span covers many CPI release cycles, which gives you room to calibrate and backtest.

Subscribe for updates

We‘ll never share your email with anyone else.

Subscribe

From sales and units to a price signal

Dividing SALES by UNITS for a brand on a given day gives an average realized price per unit. Because it uses net sales, it reflects what shoppers actually paid, not shelf prices. Rolling these daily values up to calendar months lines them up with the CPI reference period. You can then group brands into categories that map to the CPI components you care about, such as food at home. Chain-linking month-over-month changes within each brand, rather than comparing raw price levels across brands, keeps the series from being dominated by mix shifts.

UNIQUE_STORES and UNIQUE_CUSTOMERS help separate price moves from changes in distribution or shopper base. For example, you can drop brand-days with thin store coverage before computing averages.

Get in touch for a sample or trial of Odyssey Lite.