An hour after sunrise on June 28, Mike Safyan and nine co-workers from Planet Labs gathered in their San Francisco office for a pancake breakfast. It was a Sunday, so business was closed.
But this weekend morning was too momentous for Safyan and crew to stay home. Their 5-year-old start-up, which builds shoebox-sized satellites to orbit Earth and capture images of the planet, had eight "doves," as they're known, aboard an unmanned SpaceX capsule that was headed for the International Space Station.
If all went as planned, within about a month those satellites would join a fleet of 36 Planet Labs doves orbiting the Earth.
The big-screen TV in the dining room was turned on and Safyan chatted with some newer employees about past launches. The countdown was underway.
At 7:23 a.m. local time, a little over two minutes after blastoff, smoke ominously filled the screen. Things had gone awry. The Falcon 9 rocket was exploding just above NASA's Kennedy Space Center in Cape Canaveral, Florida, sending debris spewing into the Atlantic Ocean.
The mood in the dining room darkened.
"It's kind of a silent shock," said Safyan, who worked as an aerospace engineer at NASA before joining Planet Labs. "There's a moment of disbelief. Then it sinks in. You look at each other and realize, yep this is real—the launch failed."
Bumps and bruises are perfectly normal for emerging companies, particularly tech start-ups. But the rapidly evolving and highly mysterious business of space presents obstacles on a much grander scale.
For it to work, rockets have to first reach 10 times the speed of sound without breaking apart. The spacecrafts have to then survive the two-day journey to the International Space Station (ISS). SpaceX founder Elon Musk said this week that a structural failure caused the rocket explosion, though the investigation is ongoing.
In a successful launch, after the spacecraft reaches ISS, 250 miles above the Earth's surface, astronauts gather the satellites from onboard and send them into orbit, where systems on the ground establish contact and determine if the floating boxes are healthy and working.
Planet Labs' goal is an ambitious one. The company aims to eventually have enough doves in orbit to capture an updated view of the entire Earth every day.
Complicating matters is that launches are frequently postponed and sometimes catastrophic. Eight months before the recent failed SpaceX launch—itself delayed—a vessel from Orbital Sciences went up in flames just after takeoff, destroying 26 Planet Labs doves.
Musk said the next Falcon 9 launch will be several months away and delays will result in hundreds of millions of dollars in lost revenue.
But amid the rubble, there's tremendous optimism. Imagine the benefits of having a comprehensive and real-time view of every hidden corner of the planet, and what that means for understanding and acting upon climate change, monitoring shipping routes and deploying rescue missions.
"It allows for people to have greater insight into where there are problems and mitigate the problems before they become disasters," said Robbie Schingler, one of three Planet Labs co-founders, all former NASA employees.
Hedge fund managers, conservationists, government agencies and property developers are just some of the groups salivating over the potential.
Read More I spy... then sell the data
Silicon Valley venture capitalists are betting big, pouring $1.7 billion into space-related companies this year, according to CB Insights. Even if you exclude the $1 billion of that raised just by SpaceX, the market has still attracted almost twice as much money in 2015 as in the past three years combined. Planet Labs closed a $118 million round in April.
Google, of course, is firmly in the mix. The Web giant shelled out $500 million last year for satellite maker Skybox Imaging, a venture-backed start-up whose technology can bolster products like search, maps and Google Earth.
"There are going to be a number of companies built upon the back of all this fundamental technology," said Peter Hebert, co-founder of Lux Capital, a Silicon Valley venture firm that's investing heavily against that thesis. "This is a huge wave that's going to play out over decades."
In addition to Planet Labs, Lux has backed Kymeta, whose technology is unlocking wireless spectrum in space for use on Earth, and Orbital Insight, a software start-up also backed by Google Ventures that's processing images from satellite developers and selling its engine to hedge funds, other businesses and environmental groups.
The turning point for the industry came in 2012. That's when the SpaceX Dragon became the first commercial spacecraft to visit ISS, sparking the imagination of science fiction loving entrepreneurs everywhere.
Space scientists started fleeing NASA in droves, seeking out opportunities to build applications on top of these new, frequently deployed space vehicles. If SpaceX is sending cargo to resupply the ISS, what else could be sent up in the same vessels?
Think of it this way: SpaceX and to a lesser degree spacecraft companies like Orbital Sciences and a crop of Russian developers have laid the pipes for the next Internet.
Just like the personal computer paved the way for a new category of software companies or Amazon Web Services spawned scores of cloud applications, private spacecraft is the infrastructure enabling businesses that never before could have existed. Amazon.com founder Jeff Bezos even has his own space company, Blue Origin, which flew its first successful test flight in April.
Combine that with the rapid growth of cloud computing, big data analytics, the collapse in prices for electronic components in mobile devices and a thriving ecosystem of coders, and suddenly space is attainable and affordable.
According to SpaceX's website, a Falcon 9 launch will cost $61.2 million in 2016. The Falcon Heavy, which is set to take flight this year or next, will be "the most powerful operational rocket in the world by a factor of two," the site says. At $90 million per launch, it will be less than one-third the price of its closest competitor.
That price drop has enabled a stepped-up pace to launches, letting developers frequently test and improve their products. Planet Labs refers to the process as "agile aerospace."
"In the old days, when satellites were like mainframes, incredibly expensive, if you managed to get an image, you probably spent $10,000 on it," said Orbital Insight founder and CEO James Crawford, who previously spent four years at NASA. "You'd print it out, carry it around with you and be careful not to lose it or your boss would be pissed."
Crawford knew exactly when the tide started to change. From 2009 to 2012, he ran Google's book search project, with the goal of scanning 20 million tomes page by page (Google is now way past that number). For the search giant, it was all part of cataloging the world's information, but more broadly the project displayed just how fast and cheap imaging software had become.
What if, instead of books, Crawford could catalog the planet? He started Orbital Insight to find out and reeled in $10 million from venture investors to start building out algorithms for image processing.
Initially, Orbital Insight is taking advantage of the imagery created by legacy satellite makers DigitalGlobe and Airbus to show changing traffic patterns at major U.S. retail chains and selling access to investors.
The idea is that if every Home Depot, for example, in the Midwest sees a spike in traffic, while nearby Wal-Mart slips, that's information a hedge fund would gladly buy. When consumers grew concerned about the data breach at Target, investors could tell by the thinning parking lots.
As newer satellite makers like Planet Labs get more doves in orbit and continue to improve the resolution of their imagery, all sorts of new business opportunities will emerge, Crawford said. Already, he's looking to measure construction rates in China as well as the amount of crude oil stored in floating tanks around the globe.
That's just the very beginning.
"You get the ability to look at every cornfield at once and understand the yield of all of them or every tree in every forest at once and understand what's going on with deforestation," Crawford said.
Read MoreChina's entrepreneurial generation
Global Forest Watch (GFW), a project of the World Resources Institute, is eager to put this technology into action. Since its launch last year, GFW has been using imagery from government-sponsored satellites to improve visibility into forests and help nations around the world understand deforestation and its causes.
GFW is testing Orbital Insight's deep learning technology and expects to eventually gain access to Planet Labs' images, said Crystal Davis, director of the Washington-based organization.
The goal is to move from analyzing deforestation using images taken over the course of a year to being able to predict it and change behavior based on near real-time pictorial data. Businesses too want to understand the hidden supply chain behind their products and their impact on the environment.
Palm oil, for example, is a prominent ingredient in ice cream. In countries like Malaysia and Indonesia, forests are being chopped down and replaced by palm oil plantations.
"When Unilever produces Ben & Jerry's ice cream with palm oil, they need to be able to trace it all the way back to the plantation," said Davis. "Companies like Unilever have a lot of power to put pressure on suppliers."
Farmers also need to be ahead of the curve.
Existing satellite imagery, long a tool for the agriculture industry, is limited and backward-looking.
Wilbur-Ellis, a global distributor of agricultural products, signed a deal with Planet Labs in February designed to help growers more precisely manage their yield. The scope of Planet Labs' constellation should eventually let Wilbur-Ellis deliver updates every day instead of the current schedule of two to four times a month.
"If an agronomist sees a picture taken yesterday, he can do something," said Mike Wilbur, a vice president of Wilbur-Ellis in San Francisco.
Wilbur expects that by the 2016 planting season, Planet Labs will have enough doves in orbit so that he can provide his customers with a whole new set of tools.
That's great for land, but what about the three-quarters of the planet covered by water?
Weather pattern analysis, ship tracking and illegal fishing are some of the problems a start-up called Spire Global is aiming to address with its fleet of maritime satellites.
Founded in 2012, Spire is at an earlier stage than Planet Labs, having deployed only four satellites thus far. But it recently raised $40 million from investors including Promus Ventures and Bessemer Venture Partners so that it can send 20 satellites into orbit in the second half of this year and 100 by the end of 2017.
"When I think about a ship passing over the middle of the ocean, I have less data on that ship and the billions of dollars of goods on it than I do a UPS truck coming across town," said Chris Wake, Spire's head of operations. "We focused on maritime so we can start to get real-time tracking over all the world's oceans."
Wake said Spire has monthly launches scheduled as spacecraft makers increase the frequency of their deployments.
Planet Labs is sticking with its quarterly launch schedule. The company's doves debuted in space in April 2013, and even after the recent disasters, nine of its 11 launches have been successful.
Read MoreInternet from space?
Far from slowing down after the SpaceX explosion, Planet Labs is getting more aggressive. Last week, the company acquired satellite operator BlackBridge, gaining six years of imagery covering 6 billion square kilometers.
Planet Labs co-founder and CEO Will Marshall wrote a blog post announcing the deal, explaining that it allows "us to bring one of the largest commercial satellite imagery data sets to the Web."
It was his first post since June 28, the day of the SpaceX explosion. At that point, it was a more somber tone. He expressed his condolences to the SpaceX team and reminded his own 130 employees as well as its followers and customers that, "We've said it before: Space is hard."
Read MoreSpaceX's first failed launch to ISS
Marshall wasn't in the office the morning of the failed mission. He, like most of the company, was watching remotely early that Sunday. But Marshall quickly reached out to Safyan and the team on the ground.
"While it was unfortunate, we have a very solid plan in place going forward," Safyan said, echoing Marshall's sentiments.
And the day wasn't a complete loss, Safyan said. "The pancakes were delicious."