Earth Observation Microsatellites Are Overdone?

Space 2.0 is already here.

With the reusable rockets (congrats with the successful droneship landing, Mr. Musk!) and satellites, providing global Internet access, New Space Revolution is what actually happening right in the industry right now: space-related startups are no more a bunch of aerospace engineers but ambitious companies with successful track-record and world-changing plans.

And, along with some technological breakthroughs, one of the main drivers of that New Space Revolution is venture capital.

$1,8 bn. — 2015 VC investments in SpaceTech

According to CB Insight’s “Frontier Tech” report (thanks to Michael Dempsey)and recently published “Start-Up Space” report by Tauri Group $1,8 bn. have been invested in space-related startups. Compared with $686M investments in AR/VR, $450M investments in drones and around $2 bn. Internet of Things startups funding in 2015, this makes SpaceTech one of the highest VC’s bets at the moment.

But, let’s take a closer look at these SpaceTech deals.

Two deals that have formed vast majority of the SpaceTech funding in 2015 were $1 bn. SpaceX and $500M OneWeb ones. Thus, if we wouldn’t take these deals into account (as a some kind of one-offs), space-related funding in 2015 is just around $300M which is obviously not so epic.

And this makes a good segue to the main goal of this article:

To understand whether VC-backed Earth Observation (EO) microsatellite companies have some potential or this area is overdone already

Why the Earth Observation segment has been chosen as the topic — just because EO-related companies, claiming they would launch microsatellite constellations received more than half of all SpaceTech VC funding (excluding SpaceX and OneWeb deals) from 2012 to 2015 and 72% in 2015 — thus Earth Observation is the main investment thesis along space-related VCs at the moment.

Thereby, taking into account massive investments in microsatellites detailed before, some concerns regarding this segment is getting too crowded may occur.

And it really seems to be that way: Planet Earth is going to have tremendous +500 microsatellites Earth Observation infrastructure at 2020, counting only the spacecrafts, launched by startups (worth noting, “weather-related” companies such as Spire and PlanetiQ were not included, considering they are not really doing optical remote sensing).

But is there a market demand for such capacity?

Earth Observation: Where is the Market

Speaking of Earth Observation data market, very-high resolution (VHR; <1m GSD) data occupies 78% share, mainly driven by defense and government sales.

This can be easily demonstrated by Digital Globe’s (one of the key market players in this segment) sales structure analysis: after the acquisition of GeoEye, Digital Globe has become an undisputed champion in VHR data segment occupies more than 70% market share.

Thus, it can be assumed that we can get an insight into what is happening in this segment by analyzing Digital Globe’s sales structure (this company is publicity traded). And here’s what is going on: defense and government (primarily US one) sales are the main revenue stream for Digital Globe, accounting for 77% share.

Thereby, if we consider pretty pessimistic scenario and assume that the most of New Space startups wouldn’t be able to sell their data to defense customers due to their conservatism (although, a partnership between Urthecast and National Geospatial Agency — one of the key US EO data buyer, has been signed recently) and insufficient data quality (as it remains unproven) and take a step back to the market size slide:

There are 13 companies on the $350M market

Wow. Mind-blowing.

Obviously, this opportunity is definately not large enough to encourage VCs to allocate their funds to such startups.

But what makes investment opportunity in Earth Observation?

Earth Observation: Where is the Money

Down below I’ve listed some of the growth areas in the Earth Observation that I think are the most promising

Information Products and Analytics
As we’ve discussed before, there is going to be HUGE amount of geospatial data from all of these satellites expected to be launched. But it is still an open question, what to do with this data and how to make a transition from pixels to insights, and this is the direction Earth Observation industry is really moving towards.

There are some companies already in this area, Orbital Insight, Spaceknow and Descartes Labs to name a few that are doing things like prediction retailers profits by counting a number of cars on the parking lots and monitoring construction and manufacturing rates in China based on the satellite imagery analysis. But this is only the beginning and the full multibillion potential of Geospatial Big Data is just expected to unblocked. And some new applications of satellite imagery that have never been thought about will probably emerge — that is why we can expect different various platforms and APIs for analytics services developers (actually, this is what Digital Globe is doing with its GBDX platform, developed in partnership with Orbital Insight and Spaceknow, mentioned above).

Orbital Insight’s “parking lot” example

One more argument supporting this thesis is Skybox Imaging -> Terra Bella story: the biggest success story and single SpaceTech exit, Skybox Imaging have changed its business model, focusing more on image analysis and combining data from multiple sources (like UAVs and ground sensors).

That is why I can make a prediction (a bit bold one probably) that from now on no “satellite-only” company will raise any significant funding and that EO companies will pivot to more full-stack approach, providing information services along with raw imagery (e.g. what OmniEarth is actually doing already, with its OmniParcels service).

However, despite ambitious declarations it still uncertain (for me personally, probably Google Ventures and Lux Capital know best which companies to fund) what are the killer applications of such services that will provide cash flows and which customers see real value and ready to pay cash as well.

Disruptive pricing and distribution
Increased number of data sources allows to implement some innovation in satellite imagery pricing — very soon we can expect it to be sold by really small Areas of Interest (AOI), selected by end customer and pricing based on its square instead of scene-based sales.

If we take a closer look on the terms of buying satellite imagery, there is always some pretty large “minimal order” (up to 5000 sq.km) that can be purchased by the customer. This minimal order makes buying satellite imagery quite large investment (taking into account +60$ per sq.km prices for very-high-resolution data) irrelevant for small customers, such as individual farmers or shale oil producers.

And this is going to be changed radically due to availability of so many data sources detailed above that will allow to “split” imagery in a much greater extent that it is economically reasonable at the moment and compose mosaics for “inconvenient” objects like pipelines.

Such democratization of satellite imagery will unblock new B2B and B2C customers significantly smaller than traditional satellite imagery buyers, like government agencies and big commercial companies.

AOI-based pricing is accompanied with the new web-interfaces providing significantly better UX that traditional e-mail and phone calls approach. As an example of such web-interface take a look at Astro Digital’s imagery browser.

Another trend in satellite imagery distribution is APIs, although it may not to seem that way on the first gaze. But these APIs are another instrument (along with WEB-interfaces) to distribute imagery to final customers through developer apps.

Giving an example say there is some “FarmLogs-killer” startup looking for the NDVI calculation capabilities in its service. This startup can easily use satellite imagery API and integrate this data with just a few lines of code. On the flip side, imagery providers monetize their content via API.

Worth noting that these web-interfaces and APIs are developed by the companies operating satellite constellations (Planet Labs, Urthecast) or plan to do so (Astro Digital). Thus, it can be forecasted that traditional satellite imagery distributors will probably withdraw from the B2B market partially, replaced by New Space satellite imagery providers, distributing data via WEB-interfaces and APIs. Nevertheless, traditional imagery distributors will continue to lobby regional remote sensing projects and remain to be the main sales channel for B2G customers, which are the largest satellite imagery buyers, as we’ve seen before.

Thereby, in order to unlock new customers, New Space startups will move towards full-stack approach in this segment (distribution) of value chain also, complementing this strategy with disruptive pricing models.

Unique data
Despite that all of these satellites’ capabilities seem to be exactly the same at first site, there are a number of different onboard sensors configurations providing unmatched data such as:

  • Video
  • Multi / hyperspetral
  • SAR
  • GPS Radio Occultation
  • AIS

Conclusion

Summing up, the key takeaway from this research is:

It has become really difficult to navigate through the hype and Earth Observation segment is becoming really crowded.

On the flip side, revenue generating business models are not verified yet and it seems that company’s success is estimated based on the investments raised instead of actual sales.

However, there are some promising growth areas that I’ve outlined before and they would be researched in further articles.

Stay tuned!