Felix is developing a highly differentiated product that he’s been working on for the last 15 years that also has these massive business implications for his customers, so they’re willing to pay him way more. Typically, a radio is going for $100K, and they’re willing to pay him $500K. He can make a ton of margin on that, and they’re only willing to go to him. So he can take this commercial market that already exists, and enter it with a highly differentiated, highly defensible product that has extreme advantages for his customers. So he can basically take over that entire market.
That was Delian Asparouhov from the Venture Stories podcast, speaking about Felix Ejeckam’s company Akash Systems, which makes energy efficient and powerful RF Power Amplifier modules for commercial space satellites.
This is a feature of commercial markets basically unacceptable to government. Ejeckam is making huge margins on his product because it is differentiated and far outperforms the competitors. It also compensates for the extreme risk he has taken, having sunk many years into developing the innovation.
Yet if he wanted to sell directly to government, most likely the contract officer would ask for cost and pricing data. They would demand only 15 percent markup on input costs, regardless of the value being generated by the product. Akash Systems is the sole-source supplier of the RF power amplifier of that type. It isn’t clear that redacted invoices from commercial customers would be satisfactory. A team of auditors may descend upon the company and make a big deal of the margins. No one wants that kind of headache and negative press.
The cost-based pricing methods that government is accustomed to simply do not work in the intangible economy that drives on innovation. In one respect, a company might create a product that is several times more capable than the nearest competitor. That value should be reflected in the price, and the profit is a reward for innovation. Once competitors catch up and the product becomes a commodity, profits will be driven down.
In another respect, a company might drive down the cost of a product. If the government demands paying just 15 percent on the cost, then there would be no incentive to lower costs — such as putting up all the fixed investment that it took to get there. The effort would be reallocated due to opportunity cost.
Sometimes, costs can be lowered by orders of magnitude. When that happens, government should happily reward the company with super-profits so long as it is also getting a lower price. This will induce other firms to enter the market, and if they can match the leader, they will drive down profits. SpaceX is one example of a space company driving down costs. Here is another example from National Oceanic and Atmospheric Administration [NOAA]:
Trump/Pence might not have believed in climate change, but they were basically like, “we’re going to slash NOAA’s budget into a third, we’re going to take it from $1.5 billion to $500 million a year.” The NOAA directors were like, sh*t, we can’t afford our old school government space program, so we’re going to put out a spec of how much weather data we would like to collect. Low and behold, private companies started launching that were collect this data but in a much cheaper way. So now there’s three different companies that are fulfilling NOAA’s specs and they’re now collecting data for 1/1,000th the cost of what a government-made satellite could do.
A thousand-fold cost decrease is game changing, just like a thousand-fold capability increase. If innovators cannot return a portion of that for their investment, then it will either not happen at all or government will have to wait until it becomes common place in the commercial market before profits get driven down to acceptable ranges and companies would be willing to sell it to government.
I suppose it was lucky that there is another commercial space startup boom that was bringing technologies around to fulfill NOAA’s requirements. It isn’t clear that such an outstanding result could have been achieved by government sitting down, writing a requirement, and letting a cost-plus development contract that pays 15 percent profit. Unfortunately for the DOD, there is no organic commercial market for many military technologies. In other words, the outcome NOAA achieved will not happen for tanks, fighter aircraft, capital ships, munitions, and so forth.
Here is a previous post on this podcast discussing the effects of cost-plus contracting on the space industry.
I think a counterexample to the argument that there is no commercial market for military technologies is autonomy and machine learning. Control systems for land vehicles, ships, and aircraft all have civilian uses that are being tackled with reinforcement learning. Those models could potentially be modified for military hardware given the fundamental mechanics are not all that different. The challenge for the military is then finding isomorphic problems in the civilian sector with commercial solutions to adapt for their own use. More generically, everyone wants the ability to learn from data faster and cheaper. Some of the data science as a service companies out there deliver more actionable insights than an analyst at an order of magnitude less cost.