What should we learn from Climeworks?

Chickens and eggs.

Chicken

One of the most difficult questions we see climate founders grappling in the strange, weird world of 2025 is ordering. What should come first, the science, the messaging, or the company? It’s tough. Great science, uncoupled from the messaging and company needed to create and service demand, languishes in labs and peer review, never to be seen. We wrote about an example of this last week. A great business model lacking science and messaging will raise money (sometimes) and die. And great messaging, brought into the world without scientific rigor or favorable unit economics, is basically crypto. This stuff is hard, and especially in a market where climate companies are competing for capital with ruthless ex-Blackrock types layering AI into B2B SaaS.

But. There’s a tendency in climate circles to expect early-stage companies, even exorbitantly funded ones, to have all this figured out from the jump. Coral started out as our attempt to discover why there hadn’t been a wave of economically huge exits in climate, despite the obvious truism that every business is now a climate business, and to hopefully tilt the future curve in that direction. There are many reasons for this, which we’ve written about at great length, but one of them is the tendency in our industry towards self-flagellation.

Sometimes this is justified. We’ll never understand, for example, what the Northvolt CFO was smoking when he looked at a 500-1 debt-to-earnings and decided “yep, let’s keep approving expansions.” Others, not so much, which is why we’ve had a lot of trouble understanding the pile-on directed at DAC pioneer Climeworks, which announced last week the first round of layoffs in its history, amid what were (incorrectly, in our opinion) labeled as technology failures at their Iceland CDR plants. Let’s discuss.

Egg

First, the history. Climeworks was founded in 2009 by a pair of Swiss scientists, and reached the laboratory prototype stage in 2012. Since then, it’s launched successively larger DAC plants about every 3 years, culminating in the Orca and Mammoth plants in Iceland. These plants, which have listed DAC capacities of 4,000 and “tens of thousands” of annual tonnes, started operations in 2021 and 2024, respectively, and are the two largest establishments of their kind in the world.

Climeworks has raised almost $800m US across a half-dozen rounds, with ~$650m of that total coming in a 2022 mega-round. It’s also a cool business model, with split pathways for individuals, corporations, and tax-break seekers to purchase carbon removal, and transparent 3rd party verification. We went through and bought $100 / 100kg of removed carbon just to see how the flow worked, and let’s just say that we wish more companies would operate like this.

So, we sat down to write this piece and were expecting it to be a situation similar to Northvolt, one where good intentions ran into shitty execution, overextension of investment, a bit of light fraud, and other general incompetent malfeasance. That doesn’t appear to be the case. There are a few threads to pull on here.

1. Is Climeworks in trouble as a company?

   1. We have zero affiliation and zero inside knowledge, but there doesn’t seem any huge reason to think so. Yes, the company is cutting staff and cutting staff sucks. But, to be fair, tons of businesses are doing so at the moment, either through explicit reductions, or the cowardly bullshit of labeling layoffs as “return to office” mandates. Climeworks has clearly reached a new stage of iterative, slower growth with the opening of the Mammoth plant, and that’s a time when many startups begin to prioritize balance-sheet health and prudent spending. It still sucks, but fiduciary duty is often a heavy burden.

2. Is Climeworks a scam?

   1. The backlash against the company seems to center on two ideas; That the carbon removal available for purchase is on an admittedly-long 6 year time horizon, and that the two Iceland plants are operating at far below their listed capacity. We think the first item is a shitty look for Climeworks, but the company doesn’t hide it, and it’s been in the T&Cs of their purchase agreement since the product launched. Fair enough. As to the second… We’ll steal an elegant analogy from one of Climeworks’ founders; You can buy a car with a speedometer that goes up to 140mph, but the vehicle is impossible to drive at that speed with safety and any sort of practicality. The whole podcast linked above is worth a listen, with the particular quote at about 9:15. To make a long story long; No, we don’t think it’s a scam.

3. So, like, what’s happening here?

   1. Bad messaging, mostly. Dear Climeworks; WHY THE SHIT did you use the nameplate capacity of the plants as the value that would be reported in the media? We know the answer, which is that science is tough, these numbers are projections, and startup founders are charismatic sociopaths (this is a compliment) who like big numbers that sell and raise $. Still, if you want to pioneer DAC properly, maybe learn some prudence in how you talk about the still-nascent technology.

   2. Also, dear media; Please get someone on staff who can explain how scientific development works to you. Yes, Climeworks raised a lot of money, conducted layoffs, and their technology is still very early and improving. These are all true, but in our view they are correlative, not causative, and don’t make the company a scam. There are plenty of those around to yell about.

We wish Climeworks luck, and are cautiously optimistic about its future. The world needs DAC to work, and this still looks like the best hope.

In Other News

• Company we’re curious about; Inventwood. We have mixed feelings about the structural timber sector, mostly because cutting down trees is bad, but have accepted that it’s going to continue existing. If that’s the case, let’s make sure that we do it as little as possible, and a chemical process that enhances the durability of the same seems like a great idea. Fascinating technology, and we’ll be following as they roll out and gather real-world performance data.

• Story we loved; Project Blixt, a Google-X team which reopened older experiments in fusion, to run through them again with modernized equipment and imaging. It didn’t work, but it got us curious as to how many great ideas could be resurrected and brought to fruition with 2025 technology and compute. Bravo, gentlemen.

• Story we hated; Congressional Republicans are pushing to keep the US dependent on fossil fuels, using the incomprehensibly stupid premise that China is already so far ahead in development of clean energy technologies that it’s hardly worth contesting. Three quick thoughts; 1. There are 350m people in the US whose quality of life will be significantly better if their home country has this tech, 2. You’re not fooling anyone, just admit that BigOil is paying you to say this, and 3. Since when is the American way ceding cutting-edge technology to literally anyone? Fire all of this into the sun. We’ll do it ourselves.

Lastly, a question for our readers; We’re starting to explore ways to both make Coral more sustainable without a dependency on fundraising, and to pump capital into a number of larger projects we have cooking. Would anyone consider a small, monthly subscription to help us do that? We’ve promised to never paywall our content, and that promise stands. This would be entirely voluntary and additive, and come with access to a slack community where you can grill us on any subject, and also meet others from the best email list in climate tech. Reply back to this note with any thoughts on what you’d want to see for us to earn a few of your dollars each month. As always, thank-you for reading and we’ll see you soon.

Please consider forwarding this piece to a friend or tossing it on the social platform of your choice. Coral Carbon is free and always will be, and every subscriber helps us keep doing this. If you’re a climate founder, executive, or career-changer and want to work with us directly, stop by our online home at.

https://www.coralcarbon.io/