Software is Eating the World
“Software is eating the world” said Netscape co-founder Marc Andreessen, in 2011.
He meant that every industry would be transformed through software. Therefore, every company needed to become a software company.
He gave as an example a company we will talk about shortly: “the world’s largest bookseller, Amazon, is a software company – its core capability is its amazing software engine”.
An industry “eaten by software” would be one where practically every business transaction, important decision, and product or service offering would be facilitated or made in software, usually in a partially or fully-automated fashion and driven by big data. The industry leaders would strongly resemble – or be – software companies. Insurgents would be able to battle incumbents through lower barriers to entry and superior service offerings delivered using online services. The complexity of the underlying industry would be largely hidden from business end users, who would easily consume the services as and when they wanted.
None of this seems to apply to the space industry:
- The great achievements and advances in the space industry have been hardware advances. We think more readily of the rocket and hardware technology, and not the underlying software and data crunching that may support it.
- We would not describe the core capabilities of major space organizations as “amazing software engines”.
- Their product offerings are rarely available for online shopping.
- We cannot task an EO satellite from our mobile phone.
- An insurgent cannot beat an incumbent through the superiority of their web services.
- Many of the key transactions and service delivery processes involve numerous (usually highly technical and trained) human players
So software hasn’t eaten the space industry – why?
Why Hasn’t Software Eaten the Space Industry Yet?
One factor may be that the space industry has traditionally and understandably focused more on hardware technology advances – “space is hard” and launching and maintaining things in orbit is incredibly risky, expensive, and difficult. A risk-adverse industry does not want to “move fast and break things” via a faulty software patch.
Until recently, launch rates were low and missions required planning and hardware decisions made far in advance. This mean that available processing power to run space software has lagged far behind what Silicon Valley would consider the cutting edge and limited what could be used.
National security considerations and the role of government as the primary customer may also have played a role in determining what code was viable. The cost-plus contracting model may have discouraged software evolution since there was no downward pressure for development costs. And without many non-government downstream clients, the industry has remained a niche one for software development.
No doubt there are other reasons, but the result seems clear: software has not yet eaten the space industry. It remains heroic and epic, yet also human and analog.
This is changing rapidly.
The industry is subject to numerous pressures and influences including New Space business models, reduced launch costs and increased launches, and the wide availability of artificial intelligence, big data, and cloud computing technologies. The COVID pandemic and climate change impacts have begun to cause global economic, social, and political ruptures that will surely accelerate these trends and provide fertile ground for industry insurgents and innovators.
When Andreeson talked about the move to online services, he was referring to what we now familiarly call “as-a-service” offerings (the main kinds are Platform, Infrastructure, Software, and Data).
Characteristics of these as-a-Service models are:
- They are metered (pay per use).
- They are ubiquitous (they are available globally, with internet connectivity).
- They can scale up from thousands to millions of users.
- They are configurable by their users.
- The service provider can rapidly innovate and add new features.
- They use popular open software coding protocols for easy software integration.
I believe we are witnessing the rapid emergence of Space-as-a-Service, which is another way of saying “software is eating the space industry”.
I first blogged about the Space-as-a-Service concept in 2013, crediting Canadian company Urthecast as a pioneer.
I speculated that a space-as-a-service model would include the common characteristics of the others but would also:
“have to provide capabilities that the other -as-a-Service models could not, simply by virtue of delivering its capabilities from space. So, the litmus test is: what does the service platform offer that is unique to its extraterrestrial environment?”
And on that note…there have been some earth-shaking announcements recently.
Microsoft just announced Azure Space, a “new initiative that will deliver innovation to serve the mission needs of the space industry”. They are collaborating with illustrious partners like SpaceX and SES to offer improved access to satellite Earth Observation data, internet connectivity, ground station services, and more.
Examples they give are “integrating SpaceX’s ground stations with Azure networking capabilities”, “[SES using] Azure Orbital as a key platform to provide cloud-based managed services such as enhanced security, SD-WAN and other network functions”, and “[allowing] Azure Orbital customers…to use the [KSAT] platform to schedule satellite contacts and bring data to Azure for processing and storage.”
They join pioneer Amazon, which launched AWS Ground Station last year, which provides ability for any company to leverage ground station hardware – without much complexity – via online software services. Amazon recently announced the appointment of Peter Marquez as the company’s first-ever head of space policy, indicating the sector’s growing important to their organization.
Owner Jeff Bezos is visibly committed to space exploration – his company Blue Origin is the business proof, but I actually think his personal involvement in buying the Expanse television series for Prime Video is just as important because it shows that space is a real passion project of his, and therefore the sector is close to his heart.
So, to recap:
- The wealthiest person in the world personally loves the space industry; and
- The top two software development and cloud companies think it is a critical business sector and are partnering to deliver space hardware capabilities via simple a-la-carte online software services
This proves the viability – in fact, the inevitability – of space-as-a-service, and the rising importance of software development to the space sector.
More: it tells us that Microsoft and Amazon are likely to become binary stars of this new firmament.
They already exert enormous gravity on software developers, because developers in every industry are familiar with them, can build solutions quickly and easily on them, and have career incentives to sharpen their skills on these platforms.
Now that Microsoft and Amazon are starting to integrate their space services directly into their ecosystems – assisted by major space industry partners – their gravitic pull will massively increase until they become impossible for anyone in the space industry to ignore.
To extend my metaphor, Microsoft and Amazon’s gravity fields will pull and push each other around the sky. Even government space agencies will be affected by the perturbations of their cosmic dance.
And by making these software and partner ecosystems easily available to software developers and companies, they lower the barriers to entry. Software developers and other insurgents will be pulled in inexorably like meteor swarms, lured by profit and opportunity to shake things up.
Fast Forward 10 Years
The space industry is famously forward-thinking, so let’s explore a hypothetical business scenario which unfolds over the next decade:
- Today, Year One, company A and company B offer exactly competing space hardware services: satellite constellations that can monitor the Earth’s atmosphere. Company A invests heavily in software development support. Company B spends that equivalent time, energy, and money focusing on operational needs, some product enhancements, and sales and marketing efforts
- Year Three: Company A is getting some traction with developers, but only has some “toy” or prototype projects to point to. Company B is clearly ahead: their focus on anything but software development has visibly improved their bottom line.
- Year Five: Company A has an extensive network of developers and technical partners, and some working solutions and case studies. It is quick and easy to integrate their capabilities into business solutions. They are tightly integrated with the Amazon and Microsoft marketplaces. They recently landed their first major deal that was unambiguously due to the quality of their software. Company B is beginning to stall: their software is hard to use, there is very little uptake from the developer community, and they have no buzz.
- Year Seven: Company A’s open source components are widely used, understood, and taught in schools and online courses. Partners have now embedded its software directly in popular products and value-added extensions. Company A is now receiving a regular and increasing stream of OEM and subscription revenue. IT managers and tech leads are adopting their technology because its tested and proven. Company B sales have stalled and they have been firmly overtaken, not by any flaws in their core competency (their hardware and core service offering is now superior) but because they did not flourish in the software developer ecosystem and their competitor did.
- Year Ten: Company A buys Company B and integrates its offerings into the Company A Software-as-a-service platform.
The Butterfly Effect
Does this sound like a fantasy scenario?
It may be reassuring for a space industry incumbent to think that every one of their key business decisions is made in the (now virtual) meeting rooms by their execs – no doubt after long-term planning and analysis. Or that the hardware capabilities are the core differentiator and “moat” a competitor cannot cross.
In a space industry being eaten by software, this may no longer be true!
Important software development decisions are made daily at very junior levels, usually without executive or even managerial oversight.
They could be setting unexpected changes in motion.
Think of the Butterfly Effect, only the butterfly flapping its wings is the most junior developer who downloads an open source software component which accidentally becomes firmly embedded in the middle of the company’s EO service offering. Maybe the butterfly is the tech lead who picks the core data ETL pipeline suite with unexpected far-reaching consequences to a company’s competitive advantage (or disadvantage).
What Happens Next?
This is now the story of how software eats the space industry.
Amazon and Microsoft’s recent moves have made that fate inescapable.
The key point is software development-driven decisions, and the use of online services, are going to increasingly affect the competitiveness and strategic direction of everyone in the sector.
The story of how software eats the space industry will no doubt be an epic tale with ups and downs, fascinating themes, and dramatic incidents and conflicts. There may not be obvious heroes and villains, but there will be a host of characters. Everyone will be forced to play a part.
It is a “Choose Your Own Adventure” story.
Time to think about our next steps!