Last updated March 16, 2024
We’ve finally arrived. I’ve long been anticipating the SEC’s approval of their climate proposal, which was accepted on March 6th, 2024. We’ve been expecting this for awhile, and the going consensus for the last year and half was that the proposal was going to be accepted in October of 2023. Five months behind schedule? Not terrible.
Many are disappointed, however, appropriately noting that the proposal is a watered down version of the original proposal exactly two years ago. Others believe it is the first critical step toward increasingly stringent and impactful climate regulations. Both can be, and likely are, true.
However, after spending years in the climate space more broadly, and in the depths of climate data building my own carbon measurement software for CPG brands, I’ve come to believe that our expectations for carbon accounting are misguided, and likely will not drive change even under strict regulation. This is an important dynamic to understand, because it fundamentally changes the “job to be done” of carbon accounting SaaS, and as a result, our understanding of the landscape, the winners, the losers, and potential overlooked areas for new companies to emerge.
I call this “Carbon Accounting Software is Dead, Long Live Carbon Accounting Software” not because Carbon Accounting Software no longer has its place, but rather that the role it serves will diverge from the expectations (and valuations) we’ve had for it over the last 5 years. We’re going to talk through what that implies about the existing and future climate measurement and accounting software landscape, and where there might still be opportunity.
But first, let’s outline the disclosures and baseline how companies will have to act in the near future.
Here are the disclosure spark notes:
All of the above pertains to US publicly listed companies as well as non-US companies with US traded shares, and is expected to impact 63% of the US public equity market today. Currently only 45% of US public companies disclose their Scope 1 and Scope 2, so this new disclosure rule should lead to a material bump in that number.
As a reminder, under Sarbanes-Oxley (2022), CFOs are personally liable for the accuracy of their financial disclosures. Will people go after CFOs in the event of an incorrect disclosure? Probably, but that will be really hard to prosecute.
The major criticism I anticipate, and that we’re already seeing a bit of, is that Scope 3 is not part of the disclosure. If you’re reading this and you don’t know your scopes, here’s a quick primer:
Scope 1: Emissions you are responsible for due to your own burning of fossil fuels. If you have an onsite diesel backup generator, your burning of diesel in that generator is Scope 1. If you own a fleet of trucks for your company, your burning of gasoline when you drive is your Scope 1.
Scope 2: Emissions directly tied to the purchase of energy. In practical terms, this is simply your electricity spend. You didn’t burn the coal to make the electricity, but you paid someone to do it and deliver the electricity to you, so it’s in your Scope 2.
Scope 3: Everything else. The catch-all nature of the definition is part of the problem. Scope 3 are all of the emissions associated with the entire life cycle the other goods and services you buy. If you’re a burger company, your meat purchases are scope 3. But so are your paper purchases. And so is your Google advertising spend. And so is the landfill waste associated with the use of your products. As you might expect, this is where the lionshare of company emissions are. For your average business, 75%+ is in Scope 3. For your average consumer goods company, it’s 90%. For your average software company, it’s 95%+ Scope 3 (e.g. cloud infrastructure spend).
The chart below is a fairly comprehensive representation of the above.
Scope 3 is imperfect and highly generalized as very few companies have visibility into supply chains of their own suppliers, let alone the time to accurately run an in depth carbon accounting process for each of their suppliers. The concept of Scope 3 is, in a way, a response to the fact that most companies do not measure their Scope 1 or Scope 2. If every company in the world measured and reported on their Scope 1, the other scopes would not be necessary. But the reality is that those most incentivized to report their emissions are those most directly tied to everyday consumers, and thus those with minimal Scope 1 and Scope 2 but significant Scope 3 are those that have been most forthcoming about their emissions.
As you might imagine, given Scope 3 accounts for the vast majority of emissions, its exclusion from SEC disclosures is where most have objected to the projected efficacy of the new rule.
That said, the SEC isn’t the only governing body that will drive significant regulatory reporting change. In October 2023, California signed into law new climate disclosure rules.
Any company doing ~$710K in revenue in California and $1Bn in revenue globally will have to report their Scopes 1, 2, and 3 annually, beginning in 2027 (using 2026 data) and data will be housed in a public database by the state.
Over 5,000 companies today will be impacted by these regulations, and as a result, many businesses that may have avoided reporting Scope 3 under the SEC rules will nonetheless still be reporting Scope 3 under the CA rules.
Although increased regulation will absolutely lead to more reporting, carbon measurement at a company level will do little to reduce emissions unfortunately, and for no fault of the companies doing the reporting.
As a result, I believe almost all carbon accounting activities will converge on meeting government compliance rather than driving insights or operational alternatives and reductions, and as a result will significantly reduce the surface area for SaaS solutions to find differentiation, such as sector verticalization or proprietary data.
To understand this dynamic, we first need to understand why carbon measurement at the company level fails to drive emissions reductions.
In 2022 my co-founder and I launched a climate measurement platform for food & beverage CPG brands. The most frequent request and question we received when talking to customers was “how do I compare my results to my competitors or to the industry broadly?” This is a completely fair question, as the incentive for companies to reduce their emissions is to get credit for it with consumers, which usually requires some level of comparative positioning.
However, comparative positioning is impossible because there is no such market mechanism to “value” a company’s emissions.
Here, financial accounting is actually an apt analogy to help unpack this dynamic. Financial accounting is an effort in making a series of judgement calls on a regular basis within an agreed upon framework. When should you recognize revenue? How quickly do you depreciate an asset? Can you capitalize this cost or expense it? Even commonly accepted metrics like EBITDA have immense room for interpretation (remember community-adjusted EBITDA?) and live outside the scope of Generally Accepted Accounting Principles (GAAP).
Financial accounting is a best estimate representation of the performance of the activities of a business, and the reason this works is because it is the financial market that is the final arbiter and interpreter of the value of those activities, with accounting being one input. Said another way, the market may interpret two companies with similar financial results very differently. A public SaaS company growing revenue at 60% YoY with -20% EBITDA margins may be valued in the tens of billions in the public market even today, whereas a CPG brand of the same profile would not be. This public market valuation matters because it allows the company to raise additional capital, acquire businesses, and incentivize talent through equity grants, meaning businesses are incentivized to improve operational metrics that the market values most.
There is no such dynamic when it comes to carbon accounting. If Walmart emitted 100 mTCO2e last year and Snowflake emitted the same and both accountants followed ICCP guidelines, there is similarly room for interpretation in the final results, but no market to ascribe any value to those results.
As long as there is no way to incorporate carbon accounting into financial outcomes (such as through a carbon tax), carbon accounting will be an exercise in simply meeting the regulatory reporting threshold, and very little more. The incentive to rapidly reduce does not exist.
****
When we started Mammoth, we were inspired by the fact that 90% of CPG brand emissions were in Scope 3. If we could would map out a CPG brand’s operations, quickly and accurately calculate their supply chain emissions, and provide them with scenario analysis to evaluate alternative materials (i.e. PET vs recycled PET, or organic canola oil vs. conventional), we could show a clear path for these companies to drive down emissions.
Put simply, if reducing corporate emissions is really about reducing Scope 3, then there is value in measuring and modeling Scope 3 more accurately. The issue with this assumption, outside of the incentives detailed in the prior section above, is that measuring Scope 3 accurately enough to make operational decisions is a long and expensive process, and even if a company were to undergo such a process, it is unlikely that they have the capability to impose operational change in their supply chain at the level of depth required.
Let’s say your company purchases almond flour as a core ingredient in your products. Most carbon accounting processes will simply take a global emissions average from a popular database that aggregates a number of peer reviewed academic papers published over the last 20 years. This average may say that for every kg of almond flour produced, 2 kg of CO2e (carbon dioxide equivalents) are emitted. This is straightforward, but because the data source is a global average, no matter where you get your almond flour from, it will always be counted as 2kg CO2e.
A more accurate method may be to look at a different database, or at specific academic papers, to hopefully find a paper that is specific to the region from which you bought the almond flour. This is essentially what many SaaS platforms are doing when they claim to have a proprietary database, or “the largest database of emissions factors.” If you find a paper that says that almond flour produced in Thailand is actually 1.5kg CO2e / kg almond flour, then you might consider switching your supplier to Thailand. However, this conclusion is misguided. The 1.5kg is highly dependent on the specific operation studied in the paper, and switching your country of origin to Thailand has no guarantee that your new supplier will have the same operational emissions as those in the paper.
A truly accurate method would be actually conduct a study on the mill from which you purchase the almond flour as well as a study on all of the farms from which the mill purchases raw almonds so that you could precisely understand the levers driving the emissions results in your almond flour. Once you have done this enormous amount of work, you could then drive emissions reductions by working with your suppliers to change their practices. For example, work with your almond flour producer to source electricity from a renewable PPA or ask the farmers to stop using pesticides and plant cover crops, both options also being enormous and complex undertakings. Many of these supply chain changes will be more expensive, not less. Once you’ve done that, do it again for all your other major material purchases.
As you can see, this process is simply untenable. Mars, one of the largest food companies in the world, reportedly could only track down the farm source of 43% of its cocoa, let alone measure the practices of each one and impose changes.
Even with more stringent regulations, Scope 3 will be measured using the first or second methods outlined above, which does not enable measurement driven reductions. A common argument for carbon accounting is that you “can’t change what you don’t measure.” However, when it comes to Scope 3, you can rarely change even what you do measure.
Most emissions pass through a few sectors. According the EPA, US emissions are divided as follows:
Put simply, there are a handful of sectors that use fossil fuels as critical feedstocks (inputs), and the use of those feedstocks drives most emissions. It’s a fairly straightforward conclusion, but I had difficulty truly understanding it until I was working with large amounts of climate data to run carbon accounting analysis. Scope 3 data functions like a matryoshka doll of information, ultimately leading to a final, often fossil fuel based, feedstock.
Let’s take our almond flour example. Here is an oversimplified break down of all of the sources of emissions that one might account for when calculating the impact of almond flour. Highlighted in blue are the feedstocks (or outputs, like waste) that produce emissions.
https://miro.com/app/board/uXjVKfw6ej0=/?share_link_id=484626304703
Most of the emissions in this example will come from the use of electricity, natural gas, production and use of fertilizer, and the efficiency (yield) of turning almonds into flour, and the almond farm yield.
Why does all of this matter? Because when you understand where emissions come from, it becomes evident that motivating companies to, in this case, change their almond flour supplier, will not yield significant decarbonization. If we can develop and deploy cleaner electricity, electrify our machinery, and shift transportation and heavy industry (such as fertilizer and pesticide production) to use renewable feedstocks, then the diagram above begins to look a lot cleaner. But no group of companies can drive that change across such fundamental and far reaching industries in our economy, let alone companies multiple tiers removed from their operations. Decarbonization will happen at the sector level, through fundamental changes in the technologies and cost structures undergirding each sector.
So what does this all mean for the carbon measurement SaaS market? As of early 2022, there were ~100 carbon accounting software companies. Given the SEC regulation, there will likely be many more. Thus far, we’ve established a few key dynamics:
Without the ability to accurately calculate emissions, drive operational changes multiple layers deep in a supply chain, and get credit for it in the marketplace, the “job to be done” of carbon accounting merely becomes one of expending the least possible effort to be compliant with regulations.
Even in a world where the SEC regulation expands, the market for traditional carbon accounting SaaS is smaller than one might have believed only a few years ago. As a result, many of the winners have already been decided. Over the last five years, a variety of tools have launched offering specialization and verticalization, claiming to contain larger and more accurate databases. But as we’ve discussed, these points of differentiation do not solve the job to be done.
Meeting SEC regulations will be a commoditized skill. Selecting a carbon accounting tool will be about brand and product/service experience. For that reason, the behemoths of Watershed, Persefoni, Greenly, and Sweep, will continue to get bigger and suck up the capital and talent in the space. I also believe Microsoft, Salesforce, and even SAP’s carbon accounting offerings will one day become “good enough” to compete with these players. I am already seeing it happen.
Lastly, one of the areas we have not covered yet, and which likely deserves its own post, is the data transformation challenge in carbon accounting software. Data transformation is incredibly manual. To do carbon accounting properly, one must account for and categorize all of the purchases made by an enterprise. Admittedly, this is easier when limited to just Scopes 1 and 2, but this process remains manual, as companies configure their ERPs differently, and not all necessary data (such as employee transportation) lives in an ERP. The more granular the accounting gets, the more manual the process becomes. As a result, even the leading software providers are largely services businesses. As a current customer of one of the carbon accounting companies above told me, “I love it because they give me 7 people on my account, and I’m one of their smaller customers.”
As a result, the carbon accounting software market will continue to face stiff competition from not only the dominant pure play sustainability consultants such as Anthesis, Quantis, and WSP, but also smaller boutique firms that can outcompete with hands on service and lower prices. Whereas tools like Watershed and consultants like Quantis charge $150-200K annually, boutique firms can often undercut and charge $50-100K instead. Some of the Fortune 50 in the CPG world still use boutique consultants for their annual carbon accounting.
This post is not meant to be a complete refutation of the climate measurement SaaS thesis. Software does have an important role to play in climate measurement, but in a different, less traditional format than the way we view the dominant SaaS platforms today.
First and foremost, regardless of whether the new SEC rules are stringent enough, the mere presence of regulation will positively the green vortex - the growing and self-reinforcing feedback loop between policy, technology, and culture that has allowed the US to reduce emissions over the last 10 years with minimal policy intervention. Regulation will at the very least increase scrutiny of publicly traded high emitters over time, and keep decarbonization progress (or lack thereof) more in the public consciousness quantitatively than it is today.
https://ourworldindata.org/grapher/co-emissions-per-capita?tab=chart&country=~USA
But where is the opportunity for new technological solutions to drive value as well as change? If the behemoths are already locked in, is there space for anyone else?
Yes, but they will not be traditional accounting SaaS applications. They will need to tackle the market from an entirely different angle, and tackle a different but related job to be done with a different customer base.
If carbon accounting will truly become a commoditized but federally mandated service, then not only will Watershed or Persefoni be providing the service, but so will every core business system of record that wants further entrench itself in its customers’ operations. Salesforce, Microsoft, and SAP all have carbon accounting solutions. I believe every major fintech company tack this service on, whether it’s corporate cards like Ramp/Brex or neobanks like Chime or Revolut.
Remember, carbon accounting is about marrying transactional data, physical goods flows, and emissions data derived from years of academic studies, into one final assessment. Managing the emissions data can be an immense undertaking, and even today, it lives in unstandardized public and private databases. Most carbon accountants use similar databases, but those datasets are all delivered in massive excel files, or through software invented in the late 90s built for carbon accounting studies in academia.
Today, every SaaS platform and consultant is managing their own backend database of emissions factors, mixing together customer data, public databases, private databases, and whatever other data they can cobble together. This is untenable for many software providers, but with a unified api for emissions factor data, any core business software should be able to provide basic carbon accounting services. So much emission factor data is locked up in government databases or paywalled databases without standardization. To be able to embed a few lines of code and immediately have access to most of the available emission factor data in the world would provide a significant leap forward in access to carbon accounting.
This opportunity is still very early, and as a result, I’m only aware of two companies tackling it:
Climatiq - Founded in 2021 in Berlin, Climatiq’s goal is to make emissions factor data available to any software provider, whether they want to embed it in their existing products, or develop a full carbon accounting product offering. I met the Climatiq team back in 2022 when they were just getting started and Mammoth was looking for climate data partners. They only had 1 customer, but it’s been clear that they are now hitting their stride and that their thesis is playing out. Now, large system of record enterprises such as Celonis (business process automation), Kinaxis (end to end supply chain planning), Endava (software consulting), and IFS (ERP software) are leveraging Climatiq’s API.
Connect Earth - founded in 2021 in London, Connect Earth has a more niche approach than Climatiq, focusing exclusively on financial firms. Connect Earth’s unified emissions api allows banks to embed climate data into their products, allowing them to easily provide emissions data to consumers.
It is no surprise that both of these companies began in Europe, where regulation has existed for longer and demand for climate data has been stronger, but with SEC regulation moving forward, I believe more large software enterprises will seek out solutions like Climatiq and Connect Earth instead of building their own from scratch.