Interview with Dr. Niklas Naehrig, Head of Consulting & Sustainability, Wincasa
Sustainability reporting is becoming increasingly complex. Various benchmarks, standards and certificates require a wide range of data and KPIs. As Switzerland’s leading real estate service provider, Wincasa manages the portfolios of several large institutional owners that are directly affected by these challenges. As a property manager, Wincasa is at the source of its clients’ properties and therefore at the heart of their ESG performance. Wincasa’s sustainability team, led by Niklas Naehrig, is responsible for collecting and processing this data for ESG reporting in clients’ annual reports as well as for certificates and benchmarks. Novalytica supports Wincasa throughout the entire process with technical expertise in data collection and processing in order to automate the data flows as much as possible.
Niklas Naehrig, how have the requirements for you as property managers and for your customers, the owners, changed in relation with ESG reporting?
I see three main trends here: while the disclosure of sustainability was voluntary for a long time, it has become increasingly mandatory in the last two years. In Switzerland, for example, AMAS members are now required to publish certain environmentally relevant key figures. KGAST and ASIP also publish guidelines for their members that are closely aligned with the AMAS requirements. Looking abroad, the CSRD (Corporate Sustainability Reporting Directive) and ESRS (European Sustainability Reporting Standards) in the EU require companies to disclose their sustainability.
The second topic takes place within reporting. In the past, disclosure was already a value in itself. This is reflected, for example, in the AMAS coverage indicator, which shows the proportion of properties in the portfolio for which energy consumption data is collected. In the future, I see a clear development towards concrete action and transition plans. Owners will have to disclose the measures they are using to implement their strategies.
The most important point, however, concerns the data itself. The requirements for both the quality and quantity of the reported data have increased massively. Auditing by third parties also plays an important role here.
Recording consumption data plays a central role in ESG reporting. How do you do this at Wincasa?
When we talk about consumption, we first need to define what kind of consumption we are talking about. At Wincasa, we deal intensively with the categories of energy, water and waste. One major challenge here is electricity consumption in particular. Private metering infrastructures can be implemented to record this automatically. However, this is more of a luxury option. A large part of the recording is done manually via the consumption bills. This results in unstructured data in PDF documents that must first be structured. To make this process more efficient, we have developed an innovative solution together with Novalytica. On the one hand, it involves using artificial intelligence (AI) to read and structure the bills, but also technological support in checking the plausibility of the data.
How do you use AI and modern technologies to capture this unstructured data from consumption bills?
Energy providers are very often not yet ready to receive data directly from them. This is why we have developed an alternative solution with Novalytica. The existing PDF invoices or scans of invoices are read in and the relevant data is extracted from them using modern, AI-based algorithms. This process is carried out for each individual invoice. Various quality checks check the plausibility of the extracted data and thus ensure the high quality of the data basis.
There are hundreds of different energy plants in Switzerland, each with differently formatted bills. This is a major challenge, as the relevant data points are not always in the same place. This requires a technology that is both flexible and accurate. In addition to intelligent OCR scanning, which makes the text in the bills machine-readable, modern language models in particular help to find and extract the relevant data points in the bills automatically and with a high degree of accuracy.
We have now talked a lot about the collection of basic data. How are key figures, as required by standards such as AMAS, calculated from this?
The first step is to collect all relevant basic data. These must then be linked together. To calculate CO2 intensity, for example, area data and emission factors are required in addition to pure consumption data. These different data sources must be combined in order to calculate the required KPIs. In the background, all data must be properly recorded so that the corresponding emission factor can be added for each energy source and CO2 emissions can therefore be calculated for each energy source. Finally, the intensity is calculated using the area data. The data can now be prepared so that it is ready for publication in the annual report.
How do you ensure the quality of the data in this process?
Our recipe for success is a combination of technical expertise and modern technologies. Novalytica has used artificial intelligence and modern algorithms to implement various data quality checks that warn, for example, if a data point deviates more than the defined threshold from the value of the previous period or if a data value is missing. Combined with the experience of Wincasa’s sustainability team, which checks such warnings or errors and corrects them if necessary, this results in very high data quality.
In order to participate in a benchmark such as GRESB, but also for other reporting standards, other data is required in addition to consumption. This is often recorded in different systems. How do you optimize the data flows for your customers so that the required reporting is ultimately created as automatically as possible?
This is effectively a major challenge. A key element here is the data strategy. In practice, our customers often request data at very short notice, for example for GRESB. Unfortunately, having to deliver this data so quickly and without a data strategy in the background is often associated with a loss of quality. We therefore encourage our customers to develop a central strategy at an early stage. The first step is to define which benchmarks/certificates to participate in and which reporting format to use. On this basis, it is then evaluated which data is required for the corresponding benchmarks/certificates and how this data can be obtained.
The data flows can now be analyzed: Which ones can be done manually, which ones are already (partially) automated? Where can they be automated and optimized?
How does Novalytica support you in this?
Together with Novalytica, we have developed a solution that automates the above-mentioned data flows as far as possible. Relevant data from various sources is transferred to a central platform, where it is processed and checked for quality. All relevant data is thus stored in a central location, the data lake. Various end products, such as a GRESB asset spreadsheet or an interactive analysis dashboard, can then be generated from this cleanly prepared data basis at the touch of a button. Together, we combine expertise and know-how from management and modern data technologies to provide our clients with the key figures required for ESG reporting with as little effort as possible and in high quality.