25/05/2014
Google made a play for the skies in April when it swept in and acquired Titan Aerospace amid reports that the drone maker was being pursued by Facebook. But Google’s interest in hardware companies that provide primary sources of data is not ending there. Google is closing in on a deal for Skybox Imaging, a satellite company that specialises in recording very detailed landscape pictures and video, TechCrunch has heard from three different sources.
Google’s interest in Skybox, first reported in April by The Information, had already reached an advanced stage several weeks ago, with the two meeting in at least three rounds of acquisition talks, one source tells us. This was after people in the satellite industry started to hear rumors that Google was eyeing up Skybox, as well as another startup working in a similar area of satellite imaging, Planet Labs.
We now have heard that the deal with Skybox “is happening,” with one person estimating the price at roughly $1 billion. The same source said that at Skybox’s last fundraising round, when it picked up $70 million in 2012, it was valued at between $500 million and $700 million.
Google has declined to comment for this story, and Skybox did not respond to TechCrunch’s requests for comment.
So why might Google be interested in Skybox? There are a few areas where Skybox Imaging — or companies like it — could be attractive to the search giant.
The first is for more data for Google’s consumer mapping services.
Currently, Google uses a mixture of sources for Google Earth, the real-world imaging component of Google Maps. Some of those images are more up to date than others, and some are years out of date. Having its own primary source of data, updated regularly, would provide a more reliable and accurate set of data that Google could potentially expand into more consumer-focused products and services.
The second is for more data for Google’s B2B business.
Google has a business division called “Earth Enterprise” that provides mapping data for large organizations, institutions and businesses. “Google Earth Enterprise allows you to store and process terabytes of imagery, terrain and vector data on your own server infrastructure, and publish maps securely for your users to view using Google Earth desktop or mobile apps, or through your own application using the Google Maps API,” the division notes on its home page.
This seems to be a significant part of the company’s enterprise push — significant enough that news and case studies from the division feature prominently on Google’s general enterprise blog.
“A constellation of small imaging satellites, like what we already have in place today with RapidEye or what Skybox may have in the future, would give Google a very reliable, rich content source for imagery analytics and related applications,” Scott Soenen, CTO at another satellite company, BlackBridge, said. He says his company is planning for the next generation of its own small superspectral satellites that will ramp up imaging capabilities further.
In both of these cases, there are signs of another interesting trend at Google: as the company continues to mature, it is increasingly exploring what other lines of business it might tap in the future beyond its current bread and butter of search-related advertising.
That’s important not just for diversification: some believe search ads will decline as the dominant force in digital advertising prominence over the next couple of years. Maps and organizing data in the physical world are clear extensions of Google’s core search business.
A third area has to do with how Skybox has been conceived as a company. Founded in 2009 by Julian Mann, Dan Berkenstock, Ching-Yu Hu, and John Fenwick, the startup has gone some way towards figuring out how to commercialise its data, rather than provide simple data sources for others to configure as many other companies in the commercial satellite industry have done.
“Skybox is looking very far downstream, and that’s an approach where people have not been very successful,” Soenen says. In that regard, Google appears to be interested not just in data but also some of the talent that’s figuring out interesting ways of using it.
A pricey and dicey game
But! There are some caveats to the Skybox information that we have heard, which are worth noting here, as they explain a bit more about the state of play in the satellite business at the moment.
The first is about the target. We have heard that Google is interested in making a play for a satellite imaging company. But there are a number of companies out there that are using small satellites and other technology to record the earth’s terrain that could fit the bill.
They include BlackBridge, which has a business called RapidEye. This provides wide-scale images of large landscapes, covering around 5 million square kilometers of ground each day; last week it picked up $22 million in funding to expand that business with more satellites. Urthecast, meanwhile, is up there creating video streams of what’s happening down below.
Planet Labs, backed by the likes of DFJ, Yuri Milner and Founders Fund among many others, has deployed nearly 30 of its “birds” — which it calls “Doves” — to continuously record and send images of the earth. While companies like BlackBridge and Urthecast are producing data that is complementary to that of Skybox, Planet Labs is more of a direct competitor, it seems.
There are more companies looking to get in on the opportunity, too. OmniEarth this week announced a partnership with Harris Corp., Draper Laboratory and Dynetics to deploy up to 18 satellites to cover 100% of the earth once a day.
“This system will generate up to 60 petabytes of scientific quality Earth observation data annually to feed the next generation of Big Data analytics,” Lars Dyrud, president and CEO of OmniEarth, said in a statement. “Our planned satellite constellation will provide the platform for users who need high-quality analytics-friendly imagery to automatically extract commerce and environmental information and make predictions.”
Given that initially the rumors were that Google might be looking at a satellite company, possibly Planet Labs or Skybox, it could be that there is intentionally inaccurate information being spread to throw people off the trail of who Google may really be pursuing.
The second caveat is about Skybox itself.
We’ve heard from two different sources that the company, which has raised $91 million to date from a number of investors that include Khosla Ventures and Bessemer Venture Partners (disclosure: CrunchFund, founded by TechCrunch founder Michael Arrington, is also an investor), will need to raise more money soon if it doesn’t make an exit.
That’s because building and launching satellites — even the small satellites that Skybox bases its business around — is a pricey game, with the cost to make one device potentially running into the millions, without operational and launch costs for that bird factored in. (Relatively speaking, these are small numbers in the satellite industry, with the types of devices made by the likes of BlackBridge, Skybox and Planet Labs often referred to as “low cost”.)
JP Morgan is among the financiers that we’ve heard mentioned in the context of raising another round. Spreading reports that the company is in play with Google could be advantageous both to current investors looking for a good valuation for the company, as well as for the company as it looks for new backers.
And it turns out that the satellite game is not only pricey, but dicey, too.
Skybox’s existing data comes from the single bird that it currently has in orbit.
It recently put in an order for some 13 more devices to be built by Space Systems/Loral, set for a launch in 2015/2016, but more immediately it has plans to launch its second satellite by the end of June. “Think of that June date as a kind of deadline for Skybox,” one source told us. “Either for an exit, or getting that financing round in place.”
Satellite launches typically take place as “secondary passengers” on larger spacecraft, with many of launches of larger spacecraft getting partly or fully funded by the government.
In the case of Skybox, its plan had been to send its second satellite up piggybacking on the launch of a Soyuz vehicle scheduled to go out from the Baikonur Cosmodrome, which is in Kazakhstan but is fully controlled by the Russian government.
In the wake of sanctions between the U.S. and Russia over tensions in Ukraine – the U.S. State Department on April 28 said it would bar exports, possibly even retroactively, of satellites to Russia if they “contribute to Russia’s military capabilities” – there have been a lot of question marks over how launches from Baikonur would be affected.
SpaceNews reported earlier this month that a Canadian satellite, which was supposed to piggyback on the same Soyuz launch as Skybox’s bird, was barred by the Canadian government from doing so, citing the Ukrainian situation.
Skybox declined to comment to SpaceNews about the impact of the political tension between the U.S. and Russia. Skybox’s are commercial satellites, but even so, the U.S. government has included commercial launches in past restrictions, such as in the case of launches from China.
The launch situation is complicated, but not insurmountable. “India has a very successful launch program, and there are commercial options in the US such as with SpaceX,” notes Soenen from Blackbridge. “So there are a handful of other options for a low-cost satellite launch.”
The one satellite that Skybox has up “doesn’t give them much capacity,” as Soenen puts it. “They wouldn’t be able to do that until they have their full constellation up there.” But nevertheless it is producing some interesting data already. “Beautiful, just beautiful,” is how another industry source described it to me.
In other words, if Google does end up buying any satellite business, it will be but one high step, and a long-term bet, into a more data-rich, map-filled future.
Additional reporting by Jonathan Shieber
