Specification for private sector data sets

See Ducky's specifications for private sector data sets to be used by Ducky Footprints, with examples of minimum requirements and ideal data sets per sector

In general, private sector datasets that are of use to Ducky Footprints are stored as spreadsheets, where each row corresponds to an aggregation of data. This could be for example total consumption of electricity per neighborhood, or total spending on a food category per municipality. Individual data is not stored by Ducky, though it helps to know how many individuals and different types of households were included in the aggregation. 

As a general rule, a spreadsheet would generally contain at  minimum the following columns:

• Some transaction category like electricity or groceries
• Some unit like kWh or NOK
• Municipality or preferably neighborhood where the aggregated transactions originated (where the people live, not where the company or shop is located)
• The amount of people aggregated, preferably split according to additional demographic information

Energy

What is the minimum data for the energy sector?

The minimum data that we require to make a discernible difference in our footprint calculations is the average energy use per house by month and municipality, and the number of households that are used to calculate this number.

What is the ideal data for the energy sector?

Our ideal dataset allows us to see not only how much electricity use varies on a spatial and monthly basis, but also how that variation is affected by the houses generating it. As such, an ideal dataset would give us not only the per household energy output figures for a given BSU in a given month, it would also give us information on the type of buildings generating those figures. 

We work with three major building types for residential housing: detached houses, row houses, and apartments. As each of these building types use electricity with varying efficiency, knowing electrical outputs for these different building types increases the accuracy of our calculations substantially. Building size also affects our estimates for electricity use, and as such having the average living area in meters squared for each building type would be an added bonus. 

Energy sector example data for downloading

• Click here to download an Excel sheet example for the energy sector. 
• The tab entitled ‘Minimum’ shows an example of data where we have average energy use per household by month and BSU, alongside the number of households used to calculate said data. 
• The tab ‘Ideal’ shows similar data, but where we have energy use per household  type as well, as well as the number of people living in the houses as well, and the average living area (bruksareal) per building type.

Finance

What is the minimum data for the finance sector?

The minimum dataset required for this to be useful would need to include spending across subcategories on a municipality basis and an annual basis. We would still need to know the number of people that contributed to the total amount spent.

What is the ideal data for the finance sector?

Our ideal dataset would give us information on spending across a wide range of subcategories, on both a monthly basis and a BSU basis. The more detailed the categorisation, the better. COICOP codes are preferable but MCC codes are also acceptable. The BSU should relate to the area the consumer lives in, not where the good or service was purchased. We would also need to know the number of people that contributed to the overall numbers, presumably this could be provided in the form of the number of accounts used.

Finance sector example data for downloading

• Click here to download an Excel sheet example for the finance sector. 
• The tab entitled ‘Minimum’ shows an example of data where we have spending data across subcategories by year and by municipality in millions of NOK, and the number of accounts contributing to the data. 
• The tab entitled ‘Ideal’ shows spending data across the same subcategories, but this time on a monthly and BSU basis.
• A third tab, entitled ‘Subcategory list’, includes relevant COICOP subcategories that we make use of. 

Food and drink

What is the minimum data for the food and drink sector?

The minimum dataset required for this to be useful would need to include spending across subcategories on an annual basis, for all of the data provider’s stores within a single municipality.  We would still need to know the number of purchases that contributed to the total amount spent.

What is the ideal data for the food and drink sector?

Our ideal dataset would give us information on spending across a wide range of subcategories, on both a monthly basis and a spatial basis. The more detailed the categorisation, the better. Spatially, we would like to know the BSU code that the store itself is part of, and if possible the BSUs whose inhabitants make up the majority of the store’s customers. We would also need to know the total number of purchases that contributed to the overall numbers. If possible, the number of user accounts would be preferable.

Food and drink sector example data for downloading

• Click here to download an Excel sheet example for the food and drink sector. 
• The tab entitled ‘Minimum’ shows an example of data where we have spending data across subcategories by year and by municipality in millions of NOK, and the number of receipts which contributed to the data in a separate data table. 
• The tab entitled ‘Ideal’ shows spending data across the same subcategories, but this time on a monthly basis and with the BSU of the store attached. This tab also includes a third data table, which indicates the BSUs whose inhabitants make up the store’s customer base.
• A third tab, entitled ‘Subcategory list’, includes the subcategories we use for food.