Four shifts changing how Nordic buildings are run
Your grid bill, your building's legal status and the way your team gets answers are all changing at once – and the monthly energy report can't keep up.
Carolina Wachtmeister, VP Product for EG Software's Facility and Energy Management business in the Nordics, sees these shifts across the more than 70,000 facilities that EG EnerKey collects energy data from. And she sees them everywhere at once.
“The key questions around energy, data and sustainability are the same in Sweden, Denmark and Norway. Laws and regulations differ slightly, but user behaviour seldom does. And we're already seeing the shift from backward-looking reporting to real-time.”
For the past decade, energy management in Nordic buildings has been a monthly discipline: meter data arrives, someone compares it with last year, a report goes to management, and the interesting anomalies are found weeks after they happened. Four things are ending that rhythm.
1. Electricity is now priced by the hour – in both directions
Grid companies are adding a third line to the electricity bill: a charge for power, meaning how much you draw at your single busiest hour. In Sweden, your two highest hours in the month typically set the fee for the whole month, so one cold morning when heating, ventilation and car chargers all start at once can cost more than 29 well-run days save. Larger businesses have paid this way for years; it is now spreading to shops, schools and smaller offices. And although Sweden has paused a plan to make it mandatory from 2027, the regulator's new assignment is a better power charge, not its abolition. The reason is capacity: in Finland alone, data centres and electric boilers with signed grid connections are expected to add up to 8 gigawatts. Three times what every household in the country draws. Building bigger lines takes years. Getting customers to spread out their consumption is faster.
Meanwhile the energy price itself swings the other way. In Finland the wholesale price was below zero for 724 hours in 2024 and 447 hours in 2025, as new wind power in the north outruns the transmission lines south. For a portfolio on a spot-price contract that is a lever: water heating, pre-heating before opening, ventilation and vehicle charging can all move a few hours into the cheap ones - but only if someone sees it while it is happening. A monthly report shows the average. The money is in the hours, and few organisations can say today which of their buildings set a new peak last month, at what hour, or why.
2. Building rules tighten, reporting rules loosen
Two EU processes are moving in opposite directions. The revised Energy Performance of Buildings Directive requires member states to set minimum energy performance standards for non-residential buildings by 1 January 2027, with the worst-performing 16 percent of the stock renovated by 2030 and 26 percent by 2033. Meanwhile the Omnibus package, in force since March 2026, cut CSRD's reporting scope from an estimated 50 000 companies to roughly 5 000 – only those with more than 1 000 employees and €450 million in turnover.
Carolina's reading is that this separates two kinds of organizations.
“Property owners today balance many perspectives. Cost, environmental impact, tenant wellbeing, and increasingly compliance. GHG and ESG reporting is often the trigger for why people first look at an energy management system. But the real value comes from the insights and the regular interaction with data that support core processes, not just the reporting itself.”
Organizations that ran energy management for the report may now feel released from it. Organizations that ran it for the buildings will notice that the buildings themselves are about to be regulated harder. The second group needs performance data that is continuous and defensible, not a once-a-year snapshot.
3. The bottleneck is time, not data
Globally, energy efficiency has improved by about 1.3 percent a year since 2019, against the roughly 4 percent needed to meet the pledge to double progress by 2030. The International Energy Agency estimates that scaling existing AI-led optimization in buildings could save around 300 TWh of electricity a year, and names the barriers as fragmented ownership, lack of digitalization and weak incentives, not missing technology.
In Nordic portfolios the data is usually there. Hourly metering is widespread, and platforms like EG EnerKey collect it from tens of thousands of facilities. What is scarce is the energy specialist's time, and the ability of everyone else - the property manager, the controller, the sustainability lead - to get an answer without going through that specialist. Carolina says this is what led EG to integrate an AI assistant in EG EnerKey.
“Time constraints and the scarcity of energy specialists are among the core reasons we integrated AI into EG EnerKey. The feedback from early users is about ease of use. It removes the learning curve completely. You ask the question that's top of mind, and you can dig through and understand your data without knowing where anything lives. People get more out of the platform and explore more than they did before. And when we interviewed users across all of EG's products about what they want from AI, the most sought-after capability was natural language query. Ask the system a question instead of learning where things live.”
4. AI moves from reporting and analyzing to action
Gartner expects more than 40 percent of agentic AI projects to be cancelled by the end of 2027, citing cost, unclear value and weak risk controls. The same firm expects a third of enterprise applications to include agentic AI by 2028, up from under 1 percent in 2024. Both are probably right: the tools that survive will be narrow, embedded in an existing workflow, and scoped so that a human still decides. Carolina's view of the next twelve months follows that pattern.
“I expect conversational interaction with data to become the default way people ask ‘why did this site's consumption spike’, rather than digging through dashboards themselves. At the same time, AI will increasingly do the first pass on anomaly detection and GHG data reconciliation automatically. The analyst's job shifts from finding the outlier to deciding what to do about it. Less manual digging, more real-time action on what the data is already telling you.”
Further out, she sees intelligence moving toward the building itself: real-time, autonomous adjustment at the edge rather than analysis in a central system after the fact. Her definition of success a year from now in EG EnerKey is the first version of an energy management system that runs parts of the core process on its own - first-pass anomaly detection and data reconciliation handled without a person in the loop, and natural language query as an everyday tool rather than a quarterly one.
Action points to handle the shift
Get hourly data structured across the whole portfolio, including the sites that have always been someone else's problem. Capacity charges and price swings are hourly phenomena. Decide who in the organization should be able to ask questions of that data directly, and remove whatever stands between them and the answer. And when you add AI, add it inside the controls your IT and security teams have already approved, so the shortcut of exporting meter data into a public chatbot never becomes tempting.
