This growing need for farm data could increase the administrative burden on farmers. To address this burden, it is important to identify and make use of the data sources already available and to consider the methods that could be deployed to manage and process the data. For non-farmer outsiders, such as companies in the food chain and government agencies, the farm can look like a black box. It is hard for the outsider to observe what is happening on the farm in terms of key sustainability concerns, such as the use of fertilisers, pesticides, and antibiotics. However, the interactions between a farm and the outside world are well-documented [3]. On most farms, undocumented cash transactions have been replaced with bank transactions and a flow of invoices (and delivery or dispatch documents) for a couple of reasons. Firstly, suppliers of farm inputs (feed, fertiliser, fuel, pesticides, etc.) and food processors or traders of agricultural products want documentation to assist in the administration of their businesses. Secondly, most farmers now have to keep accounts for fiscal purposes (VAT, income tax) and to supply their banks with financial information as part of their financial relationship with such institutions.

Farmers have two well-known accounting methods through which to organize their data: Farm Financial Accounting which is an application of conventional business accounting and Farm Management Information System (FMIS) a software system managing the day-to-day activities of the farm [4]. Farm Financial Accounting uses financial transactions (payment data) to calculate financial statements (for tax purposes and financial management). FMIS is a form of management accounting that developed out of the field records/animal records and registers of inputs and outputs per field and farm activity (crop, type of animals) to guide operational and tactical management decisions. Farm Financial Accounts focuses on monetary flows (euro amounts) with trade partners and assets, whereas the focus of the FMIS is on volumes and product flows within the farm.

As the FMIS records volume-type data (such as the use of pesticides on certain crops) that are used in operational management, it seems at first sight attractive to use these data as a basis for environmental accounting and the calculation of sustainability indicators. There are however two problems with this approach. First of all, whereas all mid-sized and large farms are obliged to keep financial records for VAT and income tax purposes (and for securing credit from a bank), there are also a lot of farms that do not use an FMIS. It is only the largest farms that have a formal FMIS. Secondly, the FMIS is much more difficult to audit in a certification process. Although nearly all farmers will record their data honestly, mistakes and deliberate misrecordings are not unthinkable, especially if good environmental performance is rewarded with support payments and/or higher output prices (as is the case with eco-schemes, organic products, or products with environmental labels) or bad performance leads to lower payments and/or lower prices.

The integration of environmental and financial accounting makes sustainability indicators auditable [5]. Financial accounting, based on the theory of double-entry accounting, has methods to verify the completeness of its dataset. One is that in modern farming in European societies payments occur by bank transaction. The use of bank account statements guarantees that all payments have been recorded. By linking invoices to these payments, there is some assurance that no invoices have been ‘forgotten’; if some invoices were not recorded, which would indicate a lower sustainability level of the farm (e.g., on pesticides), then this would not show up as a deductible cost in VAT and income tax statements. That means that coherence between the sustainability and financial statements is an important aspect of auditing farm data. Of course, these checks are not perfect. Auditors have to check whether or not cash payments were made or if descriptions on invoices are intentionally inaccurate.

The integration of financial and environmental data also has important advantages for decision making by farmers as they are confronted with trade-offs between the financial and environmental aspects of their farm management practices. In reducing the nitrate runoff, a dairy farmer could choose to lower the nitrogen surplus by using less concentrated feed or less fertiliser. Both practices would have an effect on the farm’s N-surplus per ha as well as the farm’s production costs and revenues.

To control the administrative burden involved in sustainability data provision, it is also important to consider opportunities to reduce the labour required for data management and processing. Robotic process automation (RPA) is aimed at the automatic execution of administrative tasks that reproduce the work that humans do, comparable to robotics in manufacturing. The automation is done with the help of software robots or AI workers that are able to accurately perform repetitive tasks [6]. Applications are found in domains such as purchasing and supply management [7] and also accounting [8]. Robotic process automation (RPA) or robotic accounting is expected to change accounting and auditing significantly [9,10,11]. RPA software automates the input, processing, and output of data to streamline repetitive, mundane tasks.