Interview: Pavel Machalek co-founder, SpaceKnow Inc.
February 14, 2017
The recent announcement that Earth observation data analytics provider SpaceKnow Inc. had raised $4 million in Series A financing was remarkable for two reasons: The VC was not American, but Berlin-based BlueYard Capital; and the investment target, San Francisco- and Prague-based SpaceKnow, is not primarily a play on government demand.
SpaceKnow’s showcase products, both regulars on customer Bloomberg’s 350,000 worldwide financial terminals, are the China Satellite Manufacturing and Africa Night Lights indexes.
Both use SpaceKnow’s algorithms to tease out levels of economic activity as can be observed from low-orbiting Earth observation satellites, social media and other information sources.
SpaceKnow’s China index is often at odds with Chinese government economic analyses, giving investors and other China observers a neutral assessment of the world’s second-largest economy.
The Chinese and African products are just the start. SpaceKnow wants to provide similar indicesevery nation in the world. From a financial-industry base, the company is finding its solutions attractive for multinational construction and other industries.
More recently, the U.S. military — the world’s largest buyer of commercial Earth observation data, mainly through geospatial imagery provider DigitalGlobe — has developed a taste for SpaceKnow’s products in Africa and Asia.
Pavel Machalek, SpaceKnow’s co-founder, discussed the company’s status and ambition, which he said began when he was an employee of weather data analytics provider Climate Corp., which was purchased in 2013 by Monsanto for an estimated $1.1 billion.
First there was Climate Corp.
Yes, the idea that you could make such a successful business using analytics from remote sensing operations was what SpaceKnow grew out of after the Climate Corp. sale to Monsanto.
Then our co-founder Jerry [Javornicky] set up the Prague office and things developed into the current dual structure, with San Francisco as Forward Operating Base West, and Prague as Forward Operating Base East.
We’re doing more and more defense work and a lot of it is managed from Europe for the African and Pacific theaters. We’ll be opening offices in Washington, D.C., and in New York. We need to be in Washington to be where the security business is. Our web interface for situational awareness is mainly for defense customers.
Can you name some of your clients?
I can’t disclose a lot of them. In the defense field we have GBDX with DigitalGlobe, which is a customer of ours. They pay us to analyze their imagery and we are partnering to go after the maritime automated detection market. And we have a couple of more defense customers including the Tier 1 defense primes. We’re working with one of the top five construction companies.
How far does your $4 million in Series A funding take you toward whatever your performance metric is?
I can’t go into specific numbers. But we are experiencing 10x growth this year compared to last year and are on track to do another 10x next year. That is what you have to do with a fast-growth startup to justify these investments and keep growing.
We are basically in a situation where we cannot handle the in-bound traffic and prospective customers calling on their own. We’re experiencing rapid growth in the financing and construction verticals. We have an extremely low burn rate and high income so we are well on track to break even — which is unusual for a startup, this early. This bodes well for our future growth.
You mean break even in 2017?
A: The money we raised is not to help us to survive it’s to help us scale and get more customers. We’re already pretty much self-sufficient. We don't need further investment.
Being on 350,000 Bloomberg terminals I guess helps get the word out.
We have been very fortunate to have Bloomberg as a key partner. Of course we didn't pick an easy first target: China. Every month we release a new economic indicator, which often differ substantially from the numbers published by the Chinese government’s national bureau of statistical analysis.
We look forward expand on that by basically digitizing all the economies of the world — all 220 countries. We already report on GDP from 33 countries in Africa through the Africa Night Lights Index.
The goal is to digitize the world’s physical infrastructure — all the cars, the trucks, the buildings and factories, the ports and mines — in a consistent automated, scalable manner and track it every day — a Google for physical infrastructure.
Our products are apolitical. The goal is transparency and openness. Unfortunately some countries have a problem with that.
What’s your relationship with DigitalGlobe?
DigitalGlobe is our key partner and we highly value that relationship, which has been a cornerstone of our development. The quality of their data and repeat frequency is unsurpassed in the industry. But our products incorporate many disparate data sources.
Because DigitalGlobe is such a valued partner, some of our algorithms, like our ship detector, deploy separately on their GBDX, which we remain excited about.
You’re not a regular user of imagery from Airbus or e-Geos or ImageSat or others?
A: Our goal is to be connected to all these platforms. So far we have four or five connected inside our system and we continue to sign up more. Part of the money that we raised is also to be allocated to EO observation.
To reach our goal we need to be connected to all the major EO platforms and we’re pretty close, we’re maybe 70-80 percent of the way. In the next year or two, we’ll have access not just to the archives but also the tasking of all the low Earth orbit EO satellites.
DigitalGlobe has a lock on commercial 30-cm imagery now but others are coming in addition to the multiple providers of lower-resolution products. Nothing but good news for you.
It is and 30 cm is the gold standard. There’s nothing that surpasses it for accuracy and the ability to do a ground check for whatever you measure.
But we’re seeing a lot of demand for lower resolution data, 5 meters, even 30 meters. It depends what you want to measure. Measuring a large scale area of China for decades, you can utilize many different data sources. The supply of pixels, raw data, is increasing. We have more and more data to choose from, which is wonderful news for us.
Describe your Watchdog product.
It underpins the Bloomberg product and the web product for defense, it’s our own propriety technology that correlates data from different providers to produce an alert for our customers. The customer specifies an area to watch and any time there is an overpass by any Earth observation satellite in our system, when something changes at that location — more cars, less cars, for example — we will issue an alert.
Not just a new image. It will analyze whether this is critical. We’ll issue an intelligence alert to the subscriber saying: This coal mine in Mongolia just had another hectare of ground dug up. You should look at it. It’s essentially a dog that watches and barks about changes.
It short cuts the entire OODA loop as they refer to it in the military — observe, orient, decide, act. Construction sites, mines, factories, if you need to know before anyone else, Watchdog does that. It’s activity-based intelligence.
More and more companies doing things similar to what you do.
A: We need competition and we are blessed with a lot of companies like us but that do different things in different ways. This is a massive market and it’s growing.
So this is not a crowded field in any sense. There’s a long way to go.
You mentioned UrtheCast and DigitalGlobe. These are direct competitors.
A: Yes, but let me put it this way: The term supplier, customer, competitor and partner and distributor are all getting mixed up. Most of our customers are our suppliers, and some are distributors and some are potential competitors. What does that mean for our industry when things are developing like this?
We are seeing a once-in-a-generation shift in Earth observation into a truly liberated, customer-focused, value-providing whole. And all the pieces are being reshuffled now. It’s moving away from the RFP-driven, sell-to-government model that has been prevalent for decades.
Peter B. de Selding