Provision of a Historic Scanner Dataset

Tender description

ONS is the UK's largest independent producer of official statistics and is the recognised national statistical institute for the UK. We produce a range of statistics on the economy, population and society. Our long-term strategy is to acquire electronic point of sale (scanner) data from retailers to improve the quality of our price statistics, including the Consumer Prices Index (CPI) and Gross Domestic Product. We currently collect around 200,000 prices monthly for CPI, this involves price collectors visiting stores across the UK and manually collecting each price quote. Some prices are also collected by staff based in the ONS through online research and by telephone. We'd like to transform our data collection, and use scanner and web scraped data as alternatives. These data sources would allow us to produce more detailed and better quality economic statistics. The CPI is used by the Bank of England as the UK's inflation target, price statistics are also used to uprate rail fares, benefits, government gilts, and pensions. These statistics have an impact across the UK economy and society. Our focus is to improve the UK's economic statistics. Scanner systems used by retailers (at the point of sale) provide a rich source of data covering price paid and quantities of products purchased, that could be utilised by ONS to improve the quality and timeliness of the statistics we produce. As part of the research work required to implement these data sources, the ONS are looking to procure a historic scanner dataset that would allow for testing of new methods and systems while we continue our engagement with retailers to access their scanner data in the long-term. There are several different aspects to this historic scanner dataset that we would like to access. This includes: The essential variables specified in Annex 3. The historic scanner dataset should include at least 5 retailers, specific ones are specified in the procurement criteria. The number of retailers for which data has been provided will need to be identified. Retailers can be anonymised (although it is desirable that they are not) but the data should not be aggregated across retailers Data should cover at least a four year time period, from the latest available period It is essential that all grocery products are covered, but if possible we would also like data from the full product catalogue from each retailer. After the award of the contract, we would expect delivery of the data within four weeks.