Spire has spent the past decade selling itself as the “data refinery” of space, tiny cubesats vacuuming up AIS, ADS-B, and weather signals like a swarm of space Roombas. The pitch was always that being cheap and fast meant they didn’t need to be Inmarsat or NOAA; they’d be the API you plug into when you want to know where a fishing boat, a cargo jet, or a storm front is. In practice, it’s been more like the WeWork of satellites: a great slide deck, lots of venture buzzwords, and a constant hunt for the next round of cash.

They listed on the NYSE via SPAC in 2021, right as the SPAC bubble was bursting, and the stock chart since then looks like the ski slope at Garmisch. They promised “space-as-a-service” recurring revenue, but the market has mostly seen small-volume contracts with governments and shipping firms, not the hockey-stick SaaS growth investors hallucinated. NASA and NOAA toss them some scraps, the DoD experiments with their data, and insurance and logistics companies nibble, but nobody’s betting the farm on cubesat weather reports just yet.

Technically, Spire’s constellation is respectable: a hundred-plus satellites, mostly in LEO, constantly replenished because cubesats age like fruit. The real trick is data fusion, convincing customers that fusing AIS with weather with aircraft tracking actually gives you a decision-making edge. The challenge is that incumbents like Orbcomm, Iridium, and Maxar already have entrenched customer pipelines, while Planet hoovers up most of the Earth observation hype. Spire ends up being the “interesting adjunct” rather than the indispensable feed.

They’ve leaned hard into government work lately because surprise, surprise, sovereign budgets are more reliable than startups in need of shipping telemetry. But that also puts them in the awkward position of being a defense contractor without the scale, lobbyists, or reliability pedigree of a real prime. Not exactly Lockheed Martin, more like Lockheed Martin’s intern.

The question hanging over Spire is whether they become a niche but viable data vendor, like a Bloomberg terminal for AIS nerds, or whether they get acquired for pennies by a larger player who just wants the constellation and the engineers. The current trajectory doesn’t scream independence, unless you count “independent until the cash runs out.”