A radical proposal for how to structure EU defense procurement to maximize innovation (8 min read)
Lessons learned from Ukraine, US, EU and product management
Procurement rules can be written by governments but in reality they are written in courts when losers litigate.
This means that any procurement process must be very well-documented and along as-objective criteria as possible.
In practice this means that companies have an enormous incentive to engage before the procurement selection criteria are defined as the sale is made before.
The militaries play along with that because they really need the systems that solve their problems and not the cheapest solution.
“We write a lot about historical battles but little is headed to the bureaucratic fights in Washington to procure the systems we need, it is like the modern day battle of the Potomac with significantly less blood”.
~ General David Petraeus on getting V-shaped hulls to protect against IEDs
Europe should not replicate the US procurement system
Having spoken to people from the United States defense industry, the US defense procurement has been profoundly captured by the industry and even new startups such as Anduril and Firestorm are forced to hire former Generals and military personnel to fight their respective “battles on the Potomac”.
The US defense procurement fails in many ways.
Part of this is cultural because US Generals prefer highly capable systems over cheap systems. A mindset that comes from neglecting the economic dimensions of war and the risks of failure (to their careers) embedded in most bureaucratic organizations.
The part that is systemic though is that because the military needs systems that it can use, it will typically talk to vendors first and then design the procurement criteria together with the preferred vendor.
“If you were not the one writing the tender criteria then you can forget about winning the tender” ~ Former Chief engineer of large US defense prime
This leads to regulatory capture as private companies realize that they have to be present when the decision is made on what to buy and the best way is to hire former colleagues.
That’s obviously ignoring outright corruption, which is a well-known issue in government procurement in general.
On this part, Ukraine has not innovated but some war systems are so generic (e.g. FPV drones) that the military is not afraid of getting the wrong systems and collusion is not needed.
In any case, the US procurement system creates huge bottlenecks on time of key bureaucrats and greatly advantages existing defense companies with access to the government.
Defense is uniquely vulnerable to regulatory capture
Because procurement systems interfere with normal market dynamics (buyer picks the vendor they want) and there is a lot of lock-in in military systems as switching costs are high.
However, militaries tend to make it even worse by preferring just one system for logistics reasons.
Logistically, this makes sense. You only need to acquire one type of spare parts and your military only needs to be trained once.
It’s a vulnerability on many fronts.
When the government acts as a single buyer market (monopsony) then decides to buy just ONE system it effectively creates a monopoly on the vendor side.
This is why a Patriot missile costs 2 million USD while such a missile can probably be build with today’s technology for 300,000.
The market is effectively captured and the vendor has found a way to lock-in the government and lock-out competitors.
It’s especially dangerous because the EU doesn’t seem aware of it judging from statements by Friedrich Merz and Enrico Letta that the problem in EU is that we have too many defense systems and should just go to one.
Governments should ensure a competitive defense market by buying at least three systems for each category
In the US, former General Roosevelt instituted a rule that the US military should buy 40% of the top vendor, 30% of the second vendor and another 30% of the third vendor.
This effectively ensures a market and creates incentives for new companies to enter but also for existing companies to improve.
The rule was removed by Carter to “promote competition” and subsequently the defense market started to consolidate through mergers and now it is effectively a lethargic industry designed to exact maximum profit from the US government.
It’s only because of the changing nature of warfare and the inability of incumbents to innovate that new defense startups are wining in the US.
It also protects the military from highly effective countermeasures that can destroy a single system on the battlefield.
GPS comes to mind as a clear single point of failure and Iris2 will be no different.
One only has to look at biodiversity to understand that diversity of systems is a strength and not a failure.
There are other benefits like reducing the likelihood of court litigation over procurement as there are now three winners and one system may have a relevant advantage over the other in particular (unforeseen) situations.
Ukraine has a highly functional market for defense innovation at the brigade-level
Initially it was individual teams that were buying FPV drones so that they didn’t have to go out themselves to execute their mission (e.g. such as mine-laying).
However, this was institutionalized by allowing brigades to buy the equipment they needed with a state budget and now Ukraine has built an “Amazon for defense products” where defense companies can list their products and the brigades can just buy it and get it shipped to them.
As there are around 100 brigades, this effectively created a highly competitive market for defense.
It solved the issue of sales for defense companies (always a challenge, especially if you need to convince some staff officers to co-write procurement criteria with you) and corruption is negligible as the brigades lives depend on it.
The Ukrainian government is assigning extra credits for each type of system they destroy, which has led to a move towards buying higher quality defense products that get the job done.
Markets and competition are an incredible powerful force and should be harnessed.
However, how does one do that in peacetime?
Traditionally, NATO exercises have been done to improve training but also to understand the problems faced on the battlefield and allocate investment towards resolving them.
In the future, it doesn’t need to stay with exercises because the Ukraine war is increasingly becoming a machine-on-machine war.
Ukraine is building the ‘line of drones’ and has an historic drone-only assault. This is coming from their experimental units.
You can give armies a fixed budget, a month to prepare and then have a real drone-on-drone shootout.
It would identify problems to be solved and also allow rapid experimentation with doctrine.
The military should procure based on problems, not on prescribed solutions.
In the tech sector, the best product managers I have met always focus on defining the problem rather than on the solution.
For example, a friend of mine had a case in his company where all his clients kept asking for a companion mobile app for his product which is an energy efficiency monitoring tool that monitors power usage.
Yet, when he asked the five why’s (technique to find the real underlying problem), it turns out they only need that to know that when they install the product in a cellar of a building. They need to double-check that they connected it correctly and the server is receiving the data.
Instead of building a mobile app, he modified the product to have a led-light blink when the server is receiving data.
A solution that was many orders of magnitude easier than building a mobile app.
Customers have problems they want to solve
It’s very important that the military defines their problems they want to solve and for what price point they want it solved.
In product management, we call this “Problem space” vs. “Solution space”.
Engineers are absolutely amazing at solving problems but they are terrible at figuring out what the problems actually are. Worse, most normal people are bad at identifying the real problem they have and often speak in solutions.
“I need a missile that can hit the russians 400km out with 40kg of explosive” is a very prescriptive procurement rule.
“We need to take out S-300 launchers upto 400km away” is a much more clearly defined problem. Maybe a highly precise 1kg explosive can do it delivered by air balloons.
Problem space is hard but there is a clear framework for it.
The military should apply the “Jobs-to-be-done” framework which is used in the technology industry to build disruptive technologies in a methodological way.
For example, “we need to prevent damage to underground sea cables”
This is the problem, but since no solution exists to automatically prevent the damage another job is created “We need to catch the perpetrators when they damage the cable”. However, since no technology exists to automatically catch the perpetrators another job starts. “We need to be notified immediately when the sea cable breaks” and “We need frigates close enough to arrive in time to catch the perpetrators”.
….
Fundamentally, if there would be a system that is cheap to deploy that can secure sea cables automatically then none of the other “jobs” would be necessary.
This is how a clear understanding of your “jobs-to-be-done” can drive disruptive innovation.
It also sets the budget for the solution, patrolling sea cables is very expensive.
It would make sense to invest in special sea mines that just sink any ship nearby when a cable is cut.
Problems are often already solved but they can be solved cheaper
A great example is the Patriot battery, it’s 30 year old technology and extremely expensive. Yet, EU militaries are not considering to buy cheaper technology that doesn’t exist yet and because it’s extremely capital-intensive to develop such a system, no private investor will risk his money for that.
There are many other examples such as Tomahawk missiles can be used to take out strategic targets, yet Ukraine is doing that also but with modified planes (Skyeton).
Ukraine is innovating because it doesn’t have Tomahawks but why has no modern army invented anything better? Surely there are cheaper solutions possible.
Militaries should always have procurement outstanding for cheaper solutions to their problems
Only the military really knows what it is willing to buy and how much money it is willing to allocate towards solving its problems.
Military procurement should have open tenders for all their known problems at any point in time.
Sometimes the problem is not important and then the willingness to spend money on the solution is also low.
Opening procurement at any point in time is extremely important to eliminate market risk in a market where there is literally just one buyer.
Even if the military has existing systems, there is probably a price point/capability at which they would replace it.
Militaries should externalize their internal market demand
This gives innovative companies the certainty that there is a market and enables them to invest their own capital in creating solutions that solve the problem.
However, problems rarely come alone.
It would not be useful to have a patriot missile defense system that doesn’t integrate with the nations’ radar complex.
Likewise, it’s not useful to have a cheap drone if it cannot survive the Electronic Warfare environment it will be operating in.
Solutions can have criteria that need to be met.
Important for innovators though is to get a really clear signal on which criteria are “must-have” and which ones are “good-to-have”.
For example, for missiles it’s obvious that a larger range is preferable but the real need might be at least 100km because below that you have cheap systems already.
The reason this is important is because every criteria requires “implementation investment” and startups have limited amount of money they can invest.
You may have heard the term “Minimum-viable product”, the term is often misused but in essence it means a product that solves the problem with all the “must-have” criteria satisfied.
A clear difference between “must-have” and other criteria also avoids the problem when the military sets very expensive criteria to be met (e.g. must be undetectable by radar).
Expensive criteria often drive cost overruns in government projects.
In my experience, customers are often happy with the bicycle and further development is no longer needed. Saving the client (or startup) time and money because a bicycle is much cheaper than a car.
Militaries should use a ranking system to select the top three vendors.
This essentially tells the market exactly what the military wants to buy and how it values each capability.
It also allows for impartial selection of the top 3 vendors.
It is also a known method to promote open competition among engineers.
Kaggle is most famous for hosting online data science competitions, where anybody can try their hand at solving data science challenges.
The winners are often compensated with a prize and their solution is disclosed to the public.
It allows people to submit multiple times to understand where they are in the ranking but their solution is tested by Kaggle so they cannot game the result.
This feedback loop is what drives iterative improvement
Militaries can do exactly the same, they could submit their challenge and give a procurement contract as a reward.
They could also issue intermediate rewards to the top 5 participants in order to promote investment into promising solutions that need more development.
From a market perspective, the ability to test if their solution meets a criteria is very valuable. It is also very valuable to know how the competition is doing and what criteria they have satisfied already.
For startups, it’s unlikely they will meet all the criteria at the start but they can get going and start solving some criteria.
For investors this is very valuable because they can see which startups are progressing and then they can contact them and offer them money to continue investing.
Needless to say that the military has to conduct the testing of the criteria and assign the final score as well as provide feedback on why the criteria was met or not met.
Public benchmarks are impartial and drive investment.
A public procurement based on problems, a ranking of solutions and awarding contracts to the top three would create a very liquid and dynamic competitive market that delivers solutions to problems.
It also clearly signals to the market what the military is truly looking to buy and it gives a clear roadmap to startups and investors if they have a realistic shot at competing.
Last, it makes sure that no company gets too comfortable because the military will always have a procurement call outstanding for all their problems and a disruptive innovation can always come up out of nowhere and beat them.