When Danan Margason worked as chief product officer at the carbon management advisory firm Carbon Direct, he spent a lot of time talking to landowners looking to pursue clean energy projects on their acres of land. That’s when he noticed that many of them didn’t know how to assess the value of their own property: Would it be better to make it available for a solar farm or a wind farm? Did it have what it takes to catch the eye of a data center developer?
“They were being approached by solar folks, geothermal folks, housing developers, and carbon credit developers, and they had no idea how to think about the underlying value of their land,” he told Latitude Media.
To fix the problem, Margason founded Aarden AI, a company looking to use artificial intelligence to give landowners the tools to figure out the most profitable uses for land they want to sell or buy. The startup emerged from stealth earlier this month, backed by a $4 million funding round led by Planeteer Capital; Founders Co-op and KDX Management, among others, also participated.
Aarden AI’s software allows the owner or buyer of a land parcel to pool together a host of information, including geospatial data and documents, environmental impact assessments, legal documents, potential interconnection agreements, and market pricing. The platform makes that information searchable — thanks to a natural language model — and allows customers to create deal memos and evaluate different uses for their land.
The properties that the software is evaluating are large — the startup’s clients generally own 10,000 acres or more — so its customers are mostly large businesses and institutions invested in timberland.
To evaluate whether land is well-suited, for instance, for a solar project, the software takes into consideration data such as panel degradation rates, system size, and, most importantly, interconnection costs for the area. For a data center, it takes into account land incline and pathway to power.
“We would say, ‘This is timberland today, and it could theoretically be something else — like a carbon project — but you’re going to have the most success with timber.’ And it’s valuable for them to know that they can confidently continue to manage this timber,” Margason explained. “But oftentimes there’s also other alternative uses.”
Bridge to developers
Once alternate uses for the land have been identified, Aarden also helps landowners connect with potential developers that might be interested in leasing or acquiring it.
“We’ve built out a network of developers that we think are trustworthy actors, and care about both environmental and economic outcomes,” Margason said, noting that landowners are generally “hugely skeptical” of developers, because they’ve often felt economically excluded from a project’s success.
And developers, Margason added, have many tools to figure out where and what to build, such as NCX or Paces, but “there was no equivalent on the land owner side, and it felt a huge information asymmetry problem.”
That’s even though some land, like timberland, is a valuable asset with solid returns, and the companies that own it are often large and sophisticated institutional investors. Aarden AI’s first customers, for instance, which haven’t been disclosed yet, are large timberland owners.
At the moment, Aarden is engaging its customers through a traditional software-as-a-service model based on a subscription fee. But going forward, Margason says the company will likely evolve towards usage-based fees, with “maybe some deal participation, if we help [landowners] connect with developers.”
Meanwhile, the company is using the funds from its first one to invest in the accuracy of its models, which is now at 60% — though Margason hopes to improve them to 90%. “Now, we can help clients get a first pass view and help them focus, but then we ask them to go and do the diligence themselves,” he said. “But in an ideal world, we would actually be able to help them get it almost to the finish line…[and come] up with the final underwriting.”
Aarden said it plans to do another raise in the next 18 months.
