Will Marshall - Indexing the Earth - [Invest Like the Best, EP. 251]

Episode Summary

Executive Summary: Patrick O’Shaughnessy interviews Will Marshall, co-founder and CEO of Planet, about how cheaper rockets and radically better satellites are creating a new Earth-data industry. Marshall explains Planet’s mission to image the whole Earth daily, turn imagery into actionable analytics, and support uses from agriculture and disaster response to mapping, finance, and security.

Main Topics: The space renaissance (Priority: 10/5): Cheap launch and better satellites are unlocking huge new Earth-data markets. Planet’s origin story (Priority: 9/5): Planet began in a garage with ex-NASA scientists building low-cost satellites. Daily Earth imaging (Priority: 10/5): Planet captures frequent global imagery and builds a time series of the planet. Analytics over raw imagery (Priority: 9/5): The company is moving up the stack from pictures to higher-value information products. Applications across industries (Priority: 8/5): Agriculture, government, mapping, finance, and journalism use Planet’s data. Sovereignty, security, and ethics (Priority: 7/5): Satellite imagery is lawful globally, but Planet screens sensitive customers and uses ethics review. Public-company transition (Priority: 8/5): Planet is going public to scale sales, expand products, and widen awareness.

Key Arguments: Rockets got ~4x cheaper; satellites improved ~1,000x in cost-performance. Planet’s moat is the data archive: you cannot go back and recreate it. Agriculture use cases can lift yields 20-40% while lowering inputs. Planet’s daily imagery makes the Earth searchable like Google indexes the web. Selling analytics, not just images, should increase value and margins over time. Transparency from space can help disaster response, mapping, and even peace/security.

Data Points: satellites launched: 450 satellites - Planet has successfully built and deployed this many satellites into space. orbital fleet size: just over 100 satellites - Planet reached this scale after six to seven years of development. images for each point on Earth: about 1700 images - Planet’s archive holds this many images for every point on the landmass of Earth. scan resolution: about three by three meters - Approximate pixel size of Planet’s main Earth-imaging system. image size: 47 megapixel - Each satellite image is described at this size. daily image volume: about 3 million images per day - Planet says it captures this many images to cover the landmass. area covered daily: just over 300 million square kilometers per day - Total imaging coverage across land and some coastal/ocean territories. Earth landmass: about 150 million square kilometers - Used as the comparison baseline for Planet’s daily coverage. spectral bands: eight spectral bands - The main scanning fleet collects RGB plus five additional bands. higher-resolution fleet: about 50 centimeters - Resolution of Planet’s taskable higher-resolution satellites. tasking frequency: up to 10 or 12 times per day - Planet can revisit a requested location this often. satellites in orbit: about 3,000 satellites - Estimated operating satellites in orbit today. space debris: about 100 million pieces of debris - Marshall says this is the larger orbital hazard, mostly from old hardware. satellite altitude: about 400 kilometers up - Approximate orbit of Planet’s imaging satellites. orbit lifespan: about three years - Planet’s low-orbit satellites naturally deorbit and burn up after this period. precision agriculture yield lift: 20 to 40% - Potential yield improvement from Planet-enabled precision agriculture. input reduction: similar amounts - Fertilizer and resource use can be reduced by comparable percentages. farm coverage example: about a million farmers' fields every day - Corteva is cited as a customer using Planet data at scale. revenue: over 100 million in revenue last year - Marshall cites this as evidence Planet is ready for the public markets. satellite design iterations: 18 design build iterations - Planet has iterated its satellite design many times since launch. hardware performance gain: 10,000 X - Per-satellite per-day imagery area coverage improvement from first satellite to latest. new generation output: five times more than the previous generation - SuperDove satellites deliver this much more output at roughly the same cost. spectral band expansion: doubled the number of spectral bands - Planet added more bands to meet customer needs. radio throughput: 1.6 gigabits a second - Planet’s custom radios can transmit this rate. radio range: a thousand kilometer range - Distance over which the radio throughput is achieved. ground stations: 48 ground stations - Planet has built this number of stations around the world. biodiversity loss: 70% of life on the planet - Marshall cites this as the amount lost in the last 40 years. wild mammals decline: 82% - One component of biodiversity decline cited in the conversation. insects decline: 75% - One component of biodiversity decline cited in the conversation. freshwater fish decline: 70% - One component of biodiversity decline cited in the conversation. coral systems decline: over half - Marshall says more than half of coral systems have gone. forest decline: roughly double as much forested land 40 years ago - He says the Earth had about twice as much forested area four decades ago. global forestry project: 64 tropical countries - Planet maps deforestation with Norway funding the data.

Pivotal Quotes: "We are seeing massive shifts rockets have decreased cost about 4x." — Will Marshall: He describes the launch-cost side of the space renaissance. "We are generating and transporting fastly more data, like 10 times more than we were before, more about the Earth and transmitting that around the Earth." — Will Marshall: He explains the compounding data expansion from space infrastructure. "Indexing the Earth making it searchable with the combination of the data and machine learning that sits on top." — Will Marshall: He frames Planet’s long-term mission in internet-search terms.

Implications: Planet’s next challenge is to convert unique imagery into indispensable analytics fast enough to scale, while balancing sensitive uses and proving public-market execution.