User manufacturing as an alternative model to mass customization – and how this can become the next big trend of user-driven value creation
User manufacturing is an alternative (or supplemental) idea to mass customization, building on the notion that (some!) users are able not only to configure a good within the given solution space of a manufacturer, but also (at least partly) to develop such a solution space by their own. And then transfer their individual creations in a product.
Consider a PC: Most of us are now used to the idea to mass-customize a PC using an online configuration toolkit as, e.g., Dell offers it. Here you can just select what the manufacturer has already provided. Indeed, a main task of a configuration toolkit is to exactly ensure that a custom configuration meets the pre-developed manufacturing specs and design of the producer.
But there are also some more extreme users that really build their own, very custom PCs. They do not just configure what a manufacturer has done, but really craft very individual PCs (see the projects at pimprig.com to see what I mean). In this industry, the actual manufacturing is not too difficult, as PC architectures are modular and build to be interchangeable. But you still need some skills and dedications to do so.
Here now the idea of user manufacturing starts: I have included this within the last year or so frequently in my talks and lectures, but have not blogged too much about it yet. But this posting is the start of a series of articles to formulate this idea better:
User manufacturing (perhaps there is a better term?) is a business model were users (customers) are becoming not only co-designers, but also manufacturers, using an infrastructure provided by some specialized companies.
[Update] User manufacturing is enabled by two main technologies:
(1) Easy-to-operate design software that allows users to transfer their ideas into a design without much experience in how to operate a CAD software. eMachineshop's software is a good example for this (see below). Eric von Hippel called this tools "toolkits for user innovation": Think of mass customization configurators with a much broader solution space.
(2) Easy-to-access flexible manufacturing technology. New manufacturing technologies, first of all rapid manufacturing (e.g., laser sintering or 3D printing) enable users to transfer their ideas into concrete objects -- even of they are no pure digital products. Laser printers made publishing possible for anyone (combined with DTP software to design the stuff). Similarly, future manufacturing technology will make the manufacturing of physical goods possible for everyone.
Well, perhaps not everyone but everyone interested and involved enough with the product to invest the time in the design and manufacturing. At the beginning, user manufacturers will show lead user characteristics, i.e. users that really are ahead of a trend with regard to an application and who really hope to benefit from getting a specific product design. With a continuous improvement of tools and manufaturers, however, user manufacturing will turn mainstream.
This also allows (expert) users to set up an instant company that designs, makes and globally sells physical products could become almost as easy as starting a blog or creating an eBay store — and the repercussions would really change the way we still think about manufacturing today.
In such a world, "user-generated content" would not solely refer to media (blogs, citizen reporters, YouTube movies etc.) but to just about anything: user-generated jeans, user-generated sports cars, user-generated candy bars.
Some examples:
In the world of printed goods, user manufacturing is pretty much established: Companies like Lulu.com enable everyone to become their own publisher and provide publishing and fulfillment infrastructure that up to a few years was only of the hand of a few specialized, huge publishing houses.
eMachineshop.com is a great venture that provides full-scale manufacturing capacity to everyone. Over the internet, users can here access the entire infrastructure that before was only available for "real" manufacturwrs, or demanded complicated and transaction-cost intensive search process for local job shops. But with a very flexible toolkit at eMaschineShop, users now can design their own components and place them on diverse manufacturing outlets.
A similar idea has Big Blue Saw. The company was founded by Simon Arthur, who, as a hobby and later job, build fighting robots for Battlebots, the Robot Fighting League and other robotic sporting events. Doing this, he thought about ways to make it easier for inventors, artists, and hobbyists to create anything using modern machining technology. Big Blue Saw is the result. Its customers can upload their designs to their website. We then make these designs come to life in metal and plastic through the use of advanced robotic machining technology like waterjet cutters.
These companies are doing something really new: They provide technology that before demanded high investments and operating skills not to everyone. Well, everyone that really knows to design and assemble.
To increase the potential of user manufacturing, some other companies come in. They offer not only manufacturing, but also some supporting services. And actually provide a product, but not only components. Consider Crowdspirit. This company tries to provide everyone the capability to become the make of next ipod. Their focus is electronic manufacturing. Springwise recently reported about this idea :
What blogs, citizen journalism and YouTube have done for media, CrowdSpirit hopes to do for product development. ... How it works: Inventors submit ideas for innovative new products and contributors submit problems for inventors to work on. Members vote, define a product's specifications, and can invest money to finance development. After a first prototype has been created, selected members test and help fine-tune in cooperation with manufacturers. Once the stage of product development has been completed, contributors continue to be involved, for example by acting as a product's ambassador and promoting it to retailers, or by providing product support, like translating instruction manuals.
CrowdSpirit's primary focal point is electronics with a market price below USD 190. If all goes well, this will be followed by more expensive electronics, and other sectors as the concept develops. A selection of inventions will be launched in parallel, so that the community can work on several projects at the same time.
And now Amazon:
In an interesting article (thanks to MIT colleague Ethan Mollik for this link!), USA Today technology reporter Kevin Maney places the known activities of Amazon to let others use their infrastructure in the new light of user manufacturing:
"Point, click, make a product to sell to the world ... That's the future Amazon.com CEO Jeff Bezos hopes to set in motion with the company's new direction. If you tease out Bezos' plan, you get to a point where a high school cheerleader sitting at home with a laptop could theoretically harness computing power, design capabilities, manufacturing and distribution from around the world, and make and market a cute little pink hot rod that would compete against General Motors.
... You can rent space on Amazon's computers to run a business, or rent out its transaction capabilities to sell things and collect money, or rent pieces of its warehouses and distribution system to store and ship items — or all of the above. So, with almost no start-up costs, anyone anywhere could become a retailer. It's not just contracting with Amazon to sell your stuff, the way Target does. It's leasing pieces of Amazon to create something totally unrelated to Amazon. ...
What's new about Amazon is the leap to physical products. This might be one of those evolutionary milestones, like when the first fish crawled up on land, or Jimi Hendrix discovered feedback on his electric guitar and altered the path of rock music.
Amazon's platform will be the first to include physical distribution. "You could notify us to expect inventory from you, tell us when to pick it (from warehouse shelves), and we'll send it to any address," Bezos says. "We've spent 12 years getting good at these things, so why should somebody else have to start from scratch?"
Bezos' idea cracks open an intriguing can of worms. Why shouldn't an established manufacturer do the same, leasing out factory space and industrial design teams and its expertise the same way? Sure, there are limitations. Factories aren't as flexible as warehouses or data centers, which can handle business from just about any industry. So a manufacturer's markets would be narrower. ...
Maybe this trend would not be such bad news for GM. It has excess capacity and nearly 100 years of manufacturing expertise. If it created a carmaking platform, GM could enable the creation of dozens of new niche-market car companies, all using GM to make and distribute their designs."
As Kevin Maney observes, this model is not far afield from today's contract manufacturers in Asia, which make batches of cellphones or toys or shoes on demand for Western brands. User manufacturing would transfer this model to everyone in much smaller batches, using rapid manufacturing technologies and easy, but flexible design tools.
Just imagine what would be possible if Amazon would add to its shared online-selling and distribution capabilities some physical manufacturing capacity as, e.g., offered by e-machineshop (they do this already in the context of book printing with print-on-demand). Then we all could design, click and manufacture a product to sell to the world. Welcome to the world of user manufacturing.
Context information:
- The Elite Vintners wine customization toolkit can be interoreted in this way: This is not a real configurator (as much too complex), but more the provision of the infrastructure of a professional vinery to everyone.
- Spreadshirt, Cafepress and Zazzle enable user manufacturing within a bit more constrained solution space in the fashion industry. They allow much more than the usual t-shirt configurations.
- Tim O'Reilly characterized recently Threadless as a model of user manufacturing, but I disagree. This is crowdsourcing of design, but otherwise a more traditional (if revolutionary) business model. But Tim has a number or other good examples in his post.
- The review of the history of mass customization by Donal Reddingtion also makes this bridge from mass customization to more active users.
- And researchers of user innovation like Eric von Hippel have always noted that innovative (lead) users, who find no manufacturer that would produce their idea, turn themselves into manufacturers. Lead users, however, had to build their own manufacturing capabilities. Here is a great study by Eric with Christoph Hienerth and Clariss Baldwin about this area.
- Books: Neil A. Gershenfeld: FAB: The Coming Revolution on Your Desktop--From Personal Computers to Personal Fabrication. And in GERMAN: Andreas Neef: Vom Personal Computer zum Personal Fabricator, a book on fabbing, rapid manufacturing and new flexible manufacturing technologies.
UPDATES:
- If you live in Singapore, joint this workshop exactly to the topic on Feb. 27, 2007: http://genometri.com/DIY/
- Paul Krush reports his story of opening a user manufacturing service bureau in his new blg.
"Here, the 3D skill taught in virtual worlds are a first step, but no real help ... as I had to learn from talking to RM manufacturers and researchers, that there are still a lot of challenges in coming up with clever designs that are RM ready."
Agree. But I'm not actually thinking the 3D in virtual worlds will be the starting point for most people trying to get into RM. I could do it, but the average person isn't likely to.
What I'm thinking will happen is that we'll see traditional CAD and surfacing tools (or in some cases, polygonal modeling tools) used to create RM'd designs. However, the original data can then be repurposed for 3D inside virtual worlds for advertising purposes. I already take Pro/E files into virtual worlds via one of the triangle mesh exports (.stl, .obj, .wrl). The problem right now is that the best simulation for this secondary purpose uses a unique parametric modeling format. If it used a triangle mesh format, I'd be golden.
Posted by: csven | December 12, 2006 at 11:38 AM
Thanks for these extended comments on my post, dmi and csven.
dmi, as csvwn mentioned already, there will be new economics of manufacturing with rapid manufacturing technologies which will allow to overcome the problems of today's low volume manufacturing.
In addition, we never have to forget that the recent high-volume mass production system in reality is a small-batch-high-variety manufacturing system, that comes at a high cost of forecast complexity, inventory, discounts for unwanted products, etc. ...
If we move to an on-demand manufacturing system, there are new cost saving potentials that can counterbalance the additional cost of the very low manufacturing volumes. (This is typical "mass customization thinking").
The major point of this user manufacturing theme (or desktop manufacturing, as csven calls it quite well), is -- on the long run -- in my opinion not the manufacturing part, but the design part. csvwn characterized this quite correctly: There are plenty of new possibilities for more clever product design with RM, however, you need people being able to do so. Here, the 3D skill taught in virtual worlds are a first step, but no real help ... as I had to learn from talking to RM manufacturers and researchers, that there are still a lot of challenges in coming up with clever designs that are RM ready.
To design for this demands a totally different mindset then designing for injection molding etc ...
This is, where (a) new design tools and (b) user collaboration come into place. New design tools will enable users to come up with RM-ready design. But it will be especially users who share their designs (for money or for free, as we see it today in SL, etc.). Some expert users (lead users) will develop a suitable design, and share this with others -- using networked manufacturing capability (e.g., at Amazon or eBay) to produce this design and ship it.
Posted by: Frank Piller | December 12, 2006 at 10:43 AM
I expect that either Amazon or eBay will be the one's to really get "desktop manufacturing" under way (my preferred term atm). With eBay there's already most everything in place; including a Reputation system. With Amazon, there's the recent push to provide something like what eBay offers (along with the perception that they sell *new* goods whereas eBay is still thought of as the place for secondhand stuff). Both, however, appear to be leading the way.
Interesting to me is that both are major players in virtual worlds which perhaps signals a belief that the best way to sell tangible goods (of the non-printed, non-digital media container variety) is to do so inside a 3D space. Amazon has linked SL to its catalog so users can make purchases from inside SL; and of course the Omidyar's are investors in and users of SL. It's not at the point where it needs to be (the 3D is too primitive), but I would venture that they're learning from what's available to them now with the intent of applying those lessons to the more realistic vr's sure to follow. And the viral nature of 3D objects will make the 2D web look extraordinarily primitive imo. What can be done inside a 3D space is mindblowing.
To dmi:
"And whoever produces our niche-products will have dramatically increased setup-times with their assembly lines."
Not necessarily. Rapid-manufacturing potentially does away with much of what we have been doing.
The beauty of RM is that unlike what I do when I design a product, the average person doesn't need to know about parting lines or bypass shut-offs or draft. That's because the shapes that can be made are not limited by traditional mold solutions. And that also usually means fewer parts per product because when you can't mold a shape, you either change the shape or you break it down into more parts that *can* be. Beyond that, we'll see grown parts that contain multiple materials. For example, we'll get parts composed of both plastic and metal integrated into one form (the metal could replace wires). Consequently, assembly lines as we know them today will likely be very different from what we could see in the future.
"So in the end we have to have some bigger lot sizes than 10 or 100 or 1000 pieces, dependent on the product to keep direct costs down."
You're assuming that niche products will compete with mass products. I don't see that. It'll be a long time before RM can fabricate parts faster than injection molding or most any other process currently used. That alone would prevent desktop manufacturing from emerging.
Instead, I think of this more as an evolution of the urban vinyl toy market. Those are manufactured using traditional means (usually rotomolding) but *because* of their limited quantity fetch high prices. So based on what's already going on, I'd venture we'll see short production runs and high prices. And one doesn't have to look too far for a market. As the wealthy get ever wealthier, Christie's auction house is increasingly shocked at the winning bids. Unique sells. And RM products with shapes that have never even been seen before will definitely be unique.
Posted by: csven | December 11, 2006 at 08:04 PM
If amazon does this it will be tough competition for everybody else.
Posted by: kaye - custom pcs | December 10, 2006 at 11:10 PM
sounds pretty interesting so far that "user-generated-anything"-thing. For the business administrator that i am there is one question that occurs to me: if in future we all design and produce our own product and sell it to, lets say 10 or 100 people who rated it "ok", we might have some problems with either the retail-price or the contribution margin.
And whoever produces our niche-products will have dramatically increased setup-times with their assembly lines.
So in the end we have to have some bigger lot sizes than 10 or 100 or 1000 pieces, dependent on the product to keep direct costs down.
And finally we are from where we started, again and have maybe replaced some design- or sales-department but not having the ability to produce goods in a variable way like we wanted.
Maybe I missed something here, so is anybody out there willing to discuss this one?
Posted by: dmi | December 09, 2006 at 12:32 PM