What will it cost to comply with the EU's nature reporting standard (ESRS)?

Over the next few years, as mandated by the EU's new Corporate Sustainability Reporting Directive (CSRD), as many as 50,000 companies will be required to report on their environmental impacts under ESRS — the European Sustainability Reporting Standards. Based on estimates presented below, this could create an addressable market of $30 billion per year by 2030 for CSRD reporting software and services (beyond an estimated $15B per year for auditing and staffing). Altogether, this is roughly triple the current costs associated with climate reporting.

Note: If you haven't yet read my Dummies' Guide to CSRD, TNFD, SBTN and the Nature Action Alphabet Soup, I'd suggest reading that first. Also note, there is a major set of revisions proposed to reduce the scope of reporting as part of the 'Omnibus 1 Simplification' initiative. This would gradually phase in companies over four years, starting with large companies first. A joint statement by leading investors and companies calls to preserve the core elements of the original EU sustainability framework (PDF). The piece below attempts to establish a potential market size in 2030, so I kept to the original 2024 guidelines for this piece.

Here's the math. Per EFRAG (PDF), the administrative costs for reporting under ESRS will be approximately €740,000 on average for large listed companies in addition to €430,000 in initial investments (adjusted to 2025 euros). Spread over three years that works out to about €880,000 annually for the largest listed EU companies between 2026–2028. A big chunk of this covers auditing requirements for listed companies, which EFRAG estimates will be €4.3 billion per year. Now that's just for large EU-based companies under the mandate. There are approximately 10,000 non-EU companies and 38,000 small and medium companies that also have reporting obligations:

• 11,700 large EU corporates = €10.3 billion in reporting costs
• 10,000 large non-EU corporates = €8.8 billion (assuming same as above)
• 38,000 small & medium EU corporates = €4.2 billion (assuming 1/8th the cost)

So referencing EFRAG numbers, this would put the total at about €23.3 billion per year in reporting costs. This seems almost comically low. Costs for the 2000 companies that currently report on greenhouse gas emissions under NFRD are estimated to be about €580,000 per year (adjusted to 2025 dollars). So EFRAG would have us believe that companies now reporting on an entirely new suite of landscape-level impacts (including material impacts to freshwater, oceans, and terrestrial ecosystems) as well as increased reporting on pollutants will only cost 35% more than current climate reporting requirements??

If you read through my Alphabet Soup article, which breaks down the numerous components of ESRS, you will find this claim as astonishing as I do. ESRS is orders of magnitude more complex than stand-alone climate reporting, and I would argue will be at least doubly more expensive (excluding auditing which is a fixed cost). Also, it is a bit of a leap to believe that small and medium companies will only be spending 1/8th on reporting costs versus large listed companies. Yes, the auditing requirement is removed (reducing the cost by about 40%), but regardless of a company's size, the systems that it needs to create in order to report across all the ESRS categories is significant. For the sake of argument, I'm going to assume that after removing the auditing burden, small and medium companies will incur about one-quarter the costs of a large company (roughly double the EFRAG estimate). So here is my reality-adjusted ballpark estimate for total reporting costs under CSRD:

• 11,700 large EU corporates = €16.7 billion in reporting (about €1.4 million per company)
• 10,000 large non-EU corporates = €14.2 billion (assuming same as above)
• 38,000 small & mediums = €8.1 billion (assuming 1/4 the cost less auditing)

Altogether that adds up to about €39.0 billion per year ($45 billion US) for ESRS accounting and reporting by my estimate — or roughly 10 basis points on business revenue (aka 'turnover' as they say on the Continent). How big of a professional services market will this open up? And how about software.. can new AI products, including next generation land cover models, and advanced MRV tools help reduce those costs? Or is $45 billion underestimating the complexity involved in properly complying with ESRS?

These are some of the questions I was asking myself after attending the Nature Action Dialogues in Cambridge this spring, a conference put on by the UN Environment Programme to help companies understand and utilize data on biodiversity and ecosystem services.

I will say the tenor at the conference was somewhere between overwhelm and panic. Presentation after presentation laid out the enormous challenges involved in accounting for a vast array of potential environmental impacts that companies are now expected to disclose. The Nature Positive Initiative identified 600 metrics just for the biodiversity category (one of five ESRS categories). Another top-tier consultancy has identified 1052 metrics. For the CSRD more broadly, the EU lists 84 reporting requirements and 1200 possible data points, depending upon the sector.

All of this makes the burden of climate reporting feel quite light in comparison. Climate disclosure currently costs companies globally an estimated €13.3 billion (using the above €580,000 per company estimate applied to 23,000 companies reporting through CDP), most of which is spent documenting just three relatively straightforward greenhouse gases — carbon dioxide, methane, and nitrous oxide — across Scope 1 (on-site emissions), Scope 2 (direct energy use), and Scope 3 (supply chain emissions). A huge and growing market for carbon accounting software is in place to aid corporates in their climate reporting — the market hit $17 billion in 2023 and according to Grandview Research is expected to quadruple by 2030.

For nature reporting, the metrics are significantly more complicated, and matters are worsened by an extremely underdeveloped ecosystem of public and private data providers to support system-wide transformation across company value chains. In the climate space there are literally hundreds of organizations and research institutions that provide credible data to help (and goad) companies and governments to develop their greenhouse gas inventories (e.g. CDP, Climate Action Tracker, CERES, Climate TRACE) and hundreds more that help those entities meet their net zero targets (e.g. SBTI, Climate Action 100, Climate Arc, Climate Policy Institute, Systems Change Lab, Mission Possible).

There is no 'NDP' — akin to a Carbon Disclosure Project for nature (CDP is the central repository for greenhouse gas inventories). There is nothing like 'CNRM' — an equivalent to the Corporate Climate Responsibility Monitor for nature commitments (CCRM is an involuntary transparency initiative led by NewClimate Institute holding corporations accountable for their climate targets). There isn't really even an equivalent to an IPCC for nature. The IPCC incorporates over 100 climate models (and 23 Model Intercomparison Projects) to create a vitally important common science-based reference supporting the formal climate negotiations under UNFCCC, which also benefits the private sector. Ironically, we know more about the dynamics of our atmosphere than we do the ground we walk on…

The closest equivalent to the IPCC for nature is the IPBES initiative, which is terrific but neither funded nor mandated to provide scientific consensus or core data products. For example, there isn't one globally agreed map of biodiversity habitats to support Target 3 of the UN Biodiversity Convention (30x30) or even, for that matter, a common taxonomy of ecosystem types by which such a map would be organized.* And there is no globally agreed map of degraded lands to guide Target 2 covering restoration potential, only an overlap analysis offered at very low resolution. There is also no central utility that tracks the progress of jurisdictions towards achieving the targets of the convention, much less the major corporate entities domiciled within those jurisdictions.

Many will point to the numerous open-source land cover maps provided by research institutes like NASA and JRC. I recently reviewed the 12 leading global land cover maps and found that none of them were able to differentiate between natural and non-natural lands consistently across all landscape types. They just weren't built for that purpose, and this means corporates will have to purchase expensive high-resolution maps individually from commercial providers in order to create their environmental footprint analyses (Satelligence and Space Intelligence are two companies that offer really excellent maps, though coverage is limited largely to tropical forests in both cases).

Currently the centerpiece of global data infrastructure for nature is the Integrated Biodiversity Assessment Tool (IBAT), which provides access to three main data sets — the location of Protected Areas (WDPA), Key Biodiversity Areas (KBAs), and range maps for endangered and threatened species via the International Union for the Conservation of Nature (IUCN). These data are far from complete, and the platform is severely underfunded. I asked someone on the IBAT team when a full inventory of KBAs would be ready, and they said it would take about 10 years and $100 million to complete it. This incomplete data also comes with tight restrictions — a major problem given that KBAs are arguably the most important data set for corporate footprint analysis related to biodiversity.

One major new development in the space is the Science Based Targets Network (SBTN), which in some ways is the nature equivalent to SBTi (an initiative providing corporations with a commonly agreed framework to establish net zero climate pledges). I had a front row seat to the development of the SBTN framework as a technical reviewer for Steps 1 (Assess) and 2 (Prioritize). SBTN offers a framework to help companies make nature-positive commitments in line with CSRD and TNFD, and identifies 45 leading data tools to assist companies, divided in 6 main categories:

1. Biodiversity assessment software/web applications
2. Life Cycle Assessment (LCA) tools
3. Multi-Regional Input-Output (MRIO) analysis tools
4. Geographic Information Systems (GIS-based tools)
5. Calculated Indices (e.g. BII, EII)
6. Biodiversity-related data repositories (e.g. GBIF)

A new paper just out (Barth et al. 2025) assesses these tools by performing pair comparisons across thematic objectives, examining 11 relevant metrics concerning the state of biodiversity, including:

• ecosystem extent
• ecosystem structure
• ecosystem composition
• ecosystem function
• species biodiversity (e.g. population dynamics, richness, extinction risk)
• nature's contributions to people (i.e. ecosystem services)
• soil quality (nitrogen and phosphorus)
• water quality (nitrogen and phosphorus)
• water availability
• precipitation
• temperature

Below are pair comparisons for 22 tools supporting corporate value chain assessments (1–4 green are similar, 6–9 orange-red are dissimilar):

The one thing that pops out immediately from this analysis is the extreme heterogeneity of tools that are currently available. Only about one-quarter of the tools perform comparably with each other (scoring 1–4), and many of these are actually focused on different domains. Unlike climate change reporting, which has one single domain (the atmosphere), nature reporting has four main domains — land, freshwater, oceans, and species — and each of these have a myriad of sub-components.

It's not surprising that many companies are trying to hire PhDs and other staff to develop their own internal science teams at great cost. Megan Pilsbury, the CEO of Dunya Analytics — a SaaS platform that supports companies on their environmental reporting — has pointed out that only the largest firms can afford to do this. “Most of the rest of the companies that fall in scope for CSRD, or that even just want to be leaders in their space, don't have those kinds of resources. So they're looking for the tools that can make them more efficient. They're hiring consultants..”

Numerous firms are now starting to fill this void by delivering software and services. Some of them, for example Refinq which serves financial institutions or Clarity, are marketing AI-powered analytics. Many scientists are starting their own consulting firms, some of them focusing on specific regions or domains, for example Nature Analytics which focuses on fisheries. New companies like Kanop are offering turnkey MRV solutions, and some startups like Cecil are offering data utilities to make existing spatial data more interoperable. The NatureTech Collective is now tracking this rapidly growing market sector, but it is really early days. In the next few years, we should see an explosive growth of companies coming into the environmental reporting space to claim a piece of that $45 billion in projected ESRS expenditures.

Back to the addressable market question.. A big chunk of this market will be gobbled up by auditors (roughly $8 billion or so according to the estimates above). Perhaps another 20% will be spent by larger companies that can afford to hire FTEs and build their own nature data teams ($9 billion or so), totaling roughly $15 billion. That leaves an addressable market of $30 billion for software developers and consultants supporting ESRS compliance. A new $30x30!

Most of the rest of the companies that fall in scope for CSRD.. are looking for the tools that can make them more efficient. — Megan Pilsbury

Perhaps AI will help streamline some facets of ESRS reporting, and new scientific products and MRV tools will come online to greatly reduce data acquisition costs. But in the end, we're dealing with hundreds of novel metrics, not the few dozen associated with climate disclosure. Costs could be considerably higher. Maybe the truth is somewhere between my reality-adjusted estimate of $30B and Grandview's bullish software estimate of $70B (mid-range of $50B per year) in new spending by 2030. We shall see once the Omnibus gets sorted.

*IUCN is attempting to remedy this lack of shared definitions through their Global Ecosystem Typology initiative (GET2.0). After five years in development IUCN released a Level 3 typology (sub-biomes). It is not yet field tested and finer scale typologies still need to be developed. The goal ultimately is to define a Level 5 typology (approximately 5000 ecosystem types), derived from national ecosystem inventories (Level 6). A more simplified alternative is the Ecoregions framework, which is the most widely used taxonomy in the field of biogeography, delineating 825 discrete terrestrial units (not including Antarctica).