Open Source Hardware as an example of "Deflationary Economics"

Here's a good post by Mark Suster, a VC, on "The amazing power of deflationary economics for startups", which I think neatly explains the economic and innovation model behind 3D Robotics and most other open source hardware companies. 

It starts by reminding us of the classic disruptive business model from the Innovator's Dilemma:

In the simplest form, new startups have a product that is INFERIOR to that offered by the competition but at a dramatically lower price with the seller opting for a very thin margin on their product.

Initially their only customers are people who can get by on the reduced functionality or perhaps don’t have the money to spend on the expensive product.

Often it turns out that the market is greatly expanded by having a lower price point new entrant. And over time the new entrant attracts enough business that, as depicted in the graph above, the quality of the product slowly increases over time.

The new entrant keeps margins low but suddenly has a lot of profits due to large volumes of business.

How does the incumbent respond? Not by dropping price & quality – they don’t have an advantage there. Instead they spend more money trying to innovate on product quality and call attention to the weaknesses of the new entrants product quality.

Often major customers defect en masse to the new entrant as they realize that the huge price premium is not justified by the product differentials.

It then provides some questions to ask to see if your product/service fits this bill:

  • Does your product dramatically reduce costs in an industry with large incumbents and fat margins?
  • Can you provide a narrowly focused product to a niche of that market who will be attracted to dramatically lower costs?

In the case of UAVs, both these seem true: most companies in this space price their products like military-industrial products, not consumer electronics. They typically have the margins of defense contractors. Meanwhile, the civilian/amateur market for UAVs is much more price sensitive and has been poorly served to date, both because of the high prices and also the regulatory restrictions on closed-source vendors.

The open source entrants in this space have the opportunity to be classic disruptors: faster, cheaper and, eventually, better. Starting with the original open source autopilot, Paparazzi, they were initially dismissed as being buggy, hard to use and poorly supported. But as more teams entered the fray, the open source autopilots got dramatically better, and prices continued to fall. Meanwhile the volumes rose: the ArduPilot project, for example, has shipped nearly 10,000 autopilot boards across all its varieties: that's more than all but a handful of the biggest aerospace companies.

Sure, there's still a lot of work left to do to make the open source autopilots as robust as a milspec one, but the gap is closing fast and the open source ones sell for $200 while the milspec ones go for 20 times that or more. 

At 3D Robotics (the hardware company behind the DIY Drones store) we think about open source hardware as a "90/10" opportunity: 90% of the performance of commercial alternatives at 10% of the price.

Of course, we'd like to do even better over time--we think that the open source innovation model is not just cheaper than closed source, but can be better, too (think Linux, Firefox, Apache, etc). So maybe 110/10 is possible, or even more when you include things that open source projects are good at, such as introducing new feature quickly and creating open platforms for user innovation. 

These are still early days, but I think the civilian UAV market will someday be even bigger than the military one. If the Innovator's Dilemma holds true, bet on the little guys moving fast and cheap to eventually dominate. 

Views: 1617

Comment by MarcS on December 27, 2011 at 2:44am

Hi Chris,

I think the one problem is that amateur and civilian UAV can not be seen as one.

The amateur is very price sensitive, for sure! No question he will go for Ardupilot etc.

The civilian user is focused on his application he wants to make profit with. So he looks for the system which will bring the lowest system costs in the end. And here he also has to include his working hours and things like service and so on. In this case, some 1000$ don´t matter for the system if it brings the performance...

In the near future the civilian (commercial) regulations will probably also request for certifications of the systems, which will also be pricey and only be available from companies.

That´s where I see the problems of open source in commercial UAS... 

Lot´s of room to play and show cool stuff for amateurs, anyway :-)

Comment by DaveyWaveyBunsenBurner on December 27, 2011 at 3:25am
This reeds to me like Chris's "Long Tail" theory. It makes sense, but it also opens a lot of questions (which I will only touch on here):

1. Calculating Value. If the new offering performs 1/10th the incumbent at 1/5th the cost, it isn't neccesarily good value. If the startup comes in at 1/2 price and 1/2 the performance this may be a better strategy.

2. The impact of other ideologies (open source etc) On the competition. Does having an Opensource offering devalue the closed source offerings or not affect them?

3. Most people watch DVDs, most don't fly UAVs and have no interest to, the maximum possible market size impacts this concept.

It's an interesting article but it needs more extrapolation into the ever increasingly complex business world.

3D Robotics
Comment by Chris Anderson on December 27, 2011 at 8:41am

Davey: a couple thoughts on your good questions:

1) Yes, the deflationary pressure should eventually be felt throughout the market. As the examples in that article indicate, low-priced startups tends to lower consumer expectations of price, forcing incumbents to lower their own prices. But only the new market entrants have the ability to operate successfully at such low margins, so they tend to gain market share over incumbants. 

3) Yes, but remember that the same open source hardware model can extend outwards easily. Look around your house at all the proprietary devices with embedded processors, from your thermostat to the appliance controls. Imagine if the same thing happened to them that's now happening to UAVs. As a side project, I applied the DIY Drones model to automatic sprinkler systems with a friend, and now we have OpenSprinkler. Do enough of these kinds of things, and you do have markets as big as those for people who watch DVDs ;-)

Comment by DaveyWaveyBunsenBurner on December 27, 2011 at 9:17am

Both very good points!


Comment by Roberto Navoni on December 27, 2011 at 3:38pm

Hi Chris,

I think that the main problem of that article is the simple vision of the market. That could be good aproach for a startup only if the target is to search a venture capital that put inside money. But not a good approach for long term development of company.

So you can sell your idea , and that is possible only in the USA , not in Italy :(

The best is to have a faberless eng company not a company with a internal production .

Have internal production is a big investment in people and hardware and also in quality.

Is better to have the project in your hand and decide where invest in the world for the production. So you are more flexible.

 I think that OS is a good opportunity but is not possible based all the business only on opensource hardware production . The software is a lot better... as DVD :)



3D Robotics
Comment by Chris Anderson on December 27, 2011 at 3:46pm

Roberto: Surely Arduino itself is a great example of how this can work in Italy! They have the exact same model as us and are doing great. They do their own production, just like we do. 

We used to be fabless and I can tell you from personal experience that it doesn't work with the open source hardware model at scale. You have to buy in big quantities to get decent prices and as your SKU count mounts, that means either constantly being out of stock of products while you wait to place another big batch order or having hundreds of thousands of dollars tied up in unsold inventory. We didn't set out to run a factory, but we learned that it's the only way to serve a community this big with good in-stock status, fair prices and minimal inventory. All the other Open Source Hardware companies of any scale do the same.  

BTW, we, like all other Open Source Hardware companies I know of, have refused all venture capital investment. We're all self-funded and operate on organic growth and positive cash flow. It's the best way to keep control of our own destiny. 

Comment by Roberto Navoni on December 27, 2011 at 4:35pm

Chris : there are a lot of great example of Italian company , Arduino is a great project only for educational world for first approach to  micro electronic but cannot compete in profesional market.

I think that your work on Autopilot is a lot better and professional respect of what are doing untill now by the Arduino  Guys.

Check this this is great guys with real expertise in development advanced electronics product based on linux and android. 

The Open Source is great for learn how to do a new work :) Then when you found what's your prefered business and what is your killer application you need to use your product for sell your service . 

There are a lot of company that sell their custom hardware with a good customize opensource platform and sell and support his product. The big example of this approach is Google with Android , use his opensource - software  for develop his media and ads service. They are doing a big business on opensource platform :)

I know the problem  of big production and also the problem of unsold inventory .  But the main problem to have a production farm is the investment for startup and in people  that work inside . So your product could be more expensive respect of the same from the china for example . Why Apple production is in china and not in USA ?

With Globalization approach you can decide where you have your Research department , where you have the production and where you have head quarter :)

That's could be an advantages for some people and disavantage for other :(

I prefer doing this discussion in Italian language not in English is too complex ..;)



3D Robotics
Comment by Chris Anderson on December 27, 2011 at 4:53pm

Roberto: thank you for your clear and smart answer; I can only imagine how much better it must sound in Italian! 

The question of insource/outsource production is not a simple black-and white one. It depends on volumes, and in our experience the pendulum can swing back and forth as you grow. Here's a crude chart I drew to give a sense of that from a US vs China manufacturing perspective. Apple is at the far right in the millions (where China makes most sense), we're still in the thousands (where local manufacturing makes more sense).

Comment by DaveyWaveyBunsenBurner on December 27, 2011 at 5:47pm

Forgive a likely stupid question but the graph is for APM and supporting hardware? I'm guessing the graph is very product/client base/complexity dependent?



3D Robotics
Comment by Chris Anderson on December 27, 2011 at 5:54pm

Davey: Yes, all products are different and this just reflects the kind of electronics that I've got the most experience in. The point of the diagram was not the specifics of the numbers but to illustrate that the the insource/outsource questions is answered differently at different volume.


You need to be a member of DIY Drones to add comments!

Join DIY Drones


Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service