Chapter 5 The Long Tail of Things

Mass production works for the masses. But what works for you?

One recent Saturday, my two youngest daughters decided they wanted to redecorate their dollhouse. They’ve been playing The Sims 3, which is a videogame that’s basically a virtual dollhouse where you can make any kind of home with a dizzying array of furniture and people choices (“Sims”), and then watch them live their lives in it.

Once their “screen time” was over, they wanted to continue play­ing out the theme with their real dollhouse. This is a sign of children brought up in the digital world, where anything is possible and every­thing is available. There are hundreds of furniture options available in The Sims. Why settle for anything less in the physical world?

But things don’t always work that way in real life. Or at least not yet.

Their first instinct, of course, was to come to me and ask me to buy new furniture for them. And my own first instinct (after saying “no” and “wait for your birthday”) was to at least find out what was avail­able. I went online and quickly realized three things: (1) dollhouse furniture is expensive; (2) there is surprisingly little variety; and (3) the stuff your kids like is invariably the wrong size for your dollhouse. Sorry, girls.

At that point, to my delight, they asked if we could make the furniture ourselves. My pleasure in their DIY spirit was slightly tem­pered, however, by memories of how projects started together with kids typically end up hours later with Dad in the workshop alone cursing broken bits of wood andX-Acto knife cuts. And even if I were to persevere, a week-long process of micro-carpentry would prob­ably end up, if history is any guide, with my clumsy bit of misshapen wood ending up in the dollhouse’s attic, unable to compete with the store-bought stuff on the other floors.

But now we have a Ç-D printer, a MakerBot Thing-O-Matic, and so this quest ended differently. We went to Thingiverse, an online re­pository of Ç-D designs that people have uploaded. And there it was, just like The Sims. Every furniture type we could want, from French Renaissance to Star Trek, was available, ready for the downloading. We grabbed some exquisite Victorian chairs and couches, resized them with a click to perfectly fit our dollhouse scale, and clicked on “build.” Twenty minutes later we had our furniture. It was free, fast, and there was so much more choice than the real world, or even Ama­zon. We may never buy dollhouse furniture again.

If you’re a toy company, this story should give you chills.

As I was writing this, Kodak went into bankruptcy, a victim of the shift away from film that needed to be bought and processed to digital photography, which is free and can be printed at home on desktop inkjet printers. If you’re making cheap plastic toys today, can you see a premonition of your future in that?

Of course, physical objects are more complex than 2-D images. Right now we can only print plastic in a few colors on our MakerBot. The finish is not as good as injection-molded plastic, and we can’t print color details with nearly as fine precision as the painting ma­chines or stencils of Chinese factories.

But that’s because we’re at the dot-matrix equivalent of Ç-D print­ers. Remember them, from the 1980s? Theywere noisy, monochrome, and crude—tiny pins hitting a black ink ribbon, little more than an automated electric typewriter. But today, just a generation later, we have cheap and silent ink-jets that print in full color with resolution almost indistinguishable from professional printing.

Now fast-forward the clock a decades or two from today’s early Ç-D printers. They will be fast, silent, and able to print a wide range of materials, from plastics to wood pulp and even food. They will have multiple color cartridges, just like your inkjet, and be able to print in as many color combinations.

Disruptive by design

Transformative change happens when industries democratize, when they’re ripped from the sole domain of companies, governments, and other institutions and handed over to regular folks.

We’ve seen this picture before: it’s what happens just before mono­lithic industries fragment in the face of countless small entrants, from the music industry to newspapers. Lower the barriers to entry and the crowd pours in.

That’s the power of democratization: it puts tools in the hands of those who know best how to use them. We all have our own needs, our own expertise, our own ideas. Ifwe are all empowered to use tools to meet those needs, or modify them with our own ideas, we will col­lectively find the full range of what a tool can do.

The Internet democratized publishing, broadcasting, and commu­nications, and the consequence was a massive increase in the range of both participation and participants in everything digital—the Long Tail of Bits.

Now the same is happening to manufacturing—the Long Tail of Things.

My first book, The Long Tail, was about exactly this—the shift in culture toward niche goods—but mostly in the digital world.

For most of the past century, the natural variation and choice in products such as music, movies, and books has been hidden by the limited “carrying capacity” of the traditional distribution systems of physical stores, broadcast channels, and megaplex movie the­aters. But once these products were available online in digital mar­ketplaces with unlimited “shelf space,” for lack of a better phrase, demand followed: the monopoly of the blockbuster was over. The mass market in culture has turned into a long tail of micro-markets, as any contact with a teenager these days will confirm (we’re all indie now!).

In short: our species turns out to be a lot more diverse than our twentieth-century markets reflected. The limited store selection of our youth reflected the economic demands of retail of the day, not the true range of human taste. We are all different, with different wants and needs, and the Internet now has a place for all of them in the way that physical markets did not.

That was not exclusively digital, of course. The Internet also lengthened the tails of physical product markets for consumers. But it did so by revolutionizing distribution, not production.

For physical goods, the twentieth-century limits to choice were based on three distribution bottlenecks—you could only buy things that passed all of the three tests:

1. The products were popular enough for manufacturers to make.

2. The products were popular enough for retailers to carry.

3. The products were popular enough for you to find (via advertis­ing or prominent placement in stores near you).

As Amazon showed, the Web could help with the latter two, right out of the gate.

First, the shift to search as a discovery mechanism meant that people could find products that were not necessarily popular enough to promote the usual way in bricks-and-mortar retail.

Second, Amazon and others who used centralized distribution warehouses and, later, distributed warehousing in the form of listings for third-party merchants who handle all the fulfillment, meant that they could list many more products than any physical retailer could carry. (Like the original catalog retailers, but without the limited pages of a paper catalog sent through the mail.)

Meanwhile, eBay did the same thing for used goods, countless specialist Web retailers emerged, and eventually Google aggregated them all into the ultimate way to find anything. Today the Web has already surfaced a Long Tail of products to rival the tail of digital ones. Bottlenecks 2 and 3 above are largely removed.

What about the first bottleneck—making more variety in the first place? Well, the Web helped some there, too.

But that was just the start. Remember that the real Web revolu­tion was not that we could just buy more stuff with greater choice, but make our own stuff that others could consume. The spread of digital cameras meant an explosion of videos that YouTube could distribute, and digital desktop tools did the same for music, publishing, and soft­ware creation. Anybody could make anything, given enough talent. Access to powerful tools and a means of distribution was no longer a barrier to participation. If you had talent and drive, you could find an audience, even if you didn’t work for the right company or have the right degree.

In the Web case, the “stuff” was and is mostly creativity and expression in digital form: words, pictures, videos, and the like. It doesn’t compete with commercial goods for money, but does compete for time. A blog may not be a book, but at the end of the day, it’s just another way to entertain and inform. The greatest change of the past decade has been the shift in time people spend consuming amateur content instead of professional content. The rise of Facebook, Tumblr, Pinterest, and all the others like them is nothing less than a massive attention shift from the commercial content companies of the twen­tieth century to the amateur content companies of the twenty-first.

Now the same is happening with physical goods. The Ç-D printers and other desktop prototyping tools are the equivalent of the cam­eras and music editing tools. They allow anyone to create one-offs for their own use. As Rufus Griscom, a web entrepreneur who founded Babble.com, puts it, “this is the Renaissance of Dilletentism."

At the same time, the world’s factories are opening up, offering Web-based manufacturing as an on-demand service to anyone with a digital design and a credit card.

The academic way to put this is that global supply chains have become “scale-free,” able to serve the small as well as the large, the garage inventor and Samsung. The non-academic way to say it is this: nothing is stopping you from making anything. The people now con­trol the means of production. Or, as The Lean Startup author Eric Reis puts it, “It’s not about ownership of the means of production, anymore. It’s about rentership of the means of production.”

Such open supply chains are the mirror of Web publishing and ecommerce a decade ago. The Web, from Amazon to eBay, revealed a Long Tail of demand for niche physical goods; now the democratized tools of production are enabling a Long Tail of supply, too.

The industrial artisan

The Long Tail of Things is around you already and has been for years, just not on this scale. Take any domain where you have a deep interest and start searching online. Got a classic car, perhaps an old MG roadster? A few clicks in your browser and you’re in the domain of hyperspecialized suppliers who focus on making nothing but re­placement bonnet release cables for car models that haven’t been made for a generation. Or perhaps you’re looking for a jewelry tree on which to dangle necklaces. You may start at Crate & Barrel, but five clicks later and you’re on Etsy, buying something much cooler and more interesting (and no more expensive) from a metal artist in Texas. The barriers to variety have disappeared.

Now the rise of the “artisanal” movement and mass-scale craft­ing has created widespread demand for such specialized goods. There is, as I write, a glut of artisanal pickle makers in Brooklyn. Mean­while, the artisanal mustard market is booming here in Berkeley; even Wal-Mart now sells more than a hundred kinds of mustard, including scores of natural stoneground varieties. The local chocolate makers, such as Tcho, compete on which has the deepest, most ethi­cal supply chain. It’s one thing to say you’re “organic” and “fair trade,” but do you start with the actual beans? And buy them straight from Ghana? And know the names of some of your pickers? For people who care about such things, it’s hard to beat the artisans for sheer obsessive caring about what they do.

What’s different about these niche physical goods, created by peo­ple and communities who aren’t attempting to conform to the eco­nomic requirements of Big Manufacturing?

For starters, niche goods aimed at discriminating audiences can command higher prices. Just think of couture fashion or fine wines. Boutique products with unique qualities are polarizing—they may be just right for you, but not for others. But the people they really are for are often willing to pay more for the privilege of being so well suited. From tailored clothes to fancy restaurants, exclusivity has al­ways commanded a premium.

This is what !.materialize, a design firm, calls “the power of the unique.” In a world dominated by one-size-fits-all commod­ity goods, the way to stand out is to create products that serve indi­vidual needs, not general ones. Custom-made bikes fit better. Right now this is mostly the privilege of the rich, as such products require hand-crafting. But what if they could be produced using digital man­ufacturing where there is no cost to complexity and no penalty for short production runs?

Increasingly, when computers are running the production machines, it costs no more to make each product different. If you’ve ever received a catalog or magazine in the mail that has a personalized message for you, that’s a formerly one-size-fits-all production machine—the printing press—turned into a digital one-size-fits-one machine, using little more than a big version of the desktop inkjet printer. Likewise when you buy a cake with fancy icing from the supermarket. That icing was applied by a robot arm—it can make each cake design dif­ferent as quickly as making them all the same—personalizing it costs no more to do, yet the supermarket can charge more for it because it is perceived as more valuable. The old model of expensive custom machines that had to make the same thing in vast numbers to justify the tooling expense is fading fast.

These niche products tend to be driven bypeoples wants and needs rather than companies’ wants and needs. Of course people have to create companies to make these goods at scale, but they work hard to retain their roots. Such entrepreneurs often state that their first obli­gation is to serve their community, and to make money second. Goods made by passionate consumers-turned-entrepreneurs tend to radiate a quality that displays craftsmanship rather than mass-manufactured efficiency.

In a sense, this is just the extreme of the specialization that Adam Smith originally recognized in The Wealth of Nations as the key to an efficient market. People should do only what they do best, he said, and trade with others who make other specialized goods. No one person or town should try to do it all, since a society can do far more collectively with efficient division of labor—comparative advantage plus trade equals growth. What was good in the eighteenth century is even better in the twenty-first, now that specialists have access to global supply chains for their commodity input materials and global consumer markets for their niche output products.

Nearly thirty years ago, two MIT professors, Michael Piore and Charles Sabel, predicted this transition in a book titled The Second Industrial Divide. They argued that the mass production model that defined twentieth-century manufacturing economies (the “first in­dustrial divide” between people and production) was neither inevi­table nor the end of innovation in making things.

Under somewhat different historical conditions, firms using a combination of craft skill and flexible equipment might have played a central role in modem economic life—instead of giving way, in almost all sectors of manufacturing, to corporations based on mass production. Had this line of mechanized craft produc­tion prevailed, we might today think of manufacturing firms as linked to particular communities rather than as the independent organizations that, through mass production, seem omnipresent.¹⁹

Today, digital desktop fabrication has indeed introduced a sort of “mechanized craft production” that Piore and Sabel could only have dreamed of. Rather than returning to the sewing machines and local machine shops that big factories drove out of the market a hundred years ago, the modern Maker Movement is built on high-tech digital fabrication, and can let regular people harness big factories at will to make what they want. It’s the perfect combination of inventing locally and producing globally, serving niche markets defined by taste, not geography. And what’s clear about these new producers is that they’re not going to be making the same one-size-fits-all products that de­fined the mass production era. Instead, they’re going be starting with one-size-fits-one and building from there, finding out how many other consumers share their interests, passions, and unique needs.

Happiness economics

What’s interesting is that such hyperspecialization is not neces­sarily a profit-maximizing strategy. Instead, it is better seen as 777∞7ZOT_tg∙-maximizing. Writing in The New York Times Magazine, Adam Davidson sees this as a natural evolution of an affluent country where the basic needs for the middle class and above have all been more than met:

The hot field ofhappiness economics argues, rather persuasively, that once people reach some level of comfort, they are willing— even eager—to trade in potential earnings at a lucrative but un­inspiring job for less (but comfortable) pay at more satisfying work. Some research by the Chicago economist Erik Hurst sug­gests that half of entrepreneurs start businesses as much to pur­sue happiness as to make money.²⁰

What’s more, consumers tend to value more highly products of which they feel they have had a hand in their creation, whether as­sembling a kit or just encouraging the creators themselves online. Re­searchers call this “the IKEA Effect,” and it dates all the way back to the Home Economics movement. As Duke University behavioral economist Dan Ariely and his colleagues write in a paper on this,

When instant cake mixes were introduced in the 1950s as part of a broader trend to simplify the life of the American housewife by minimizing manual labor, housewives were initially resistant: the mixes made cooking too easy, making their labor and skill seem undervalued. As a result, manufacturers changed the recipe to require adding an egg; while there are likely several reasons why this change led to greater subsequent adoption, infusing the task with labor appeared to be a crucial ingredient.²¹

Today, in experiments with IKEA furniture, when the paper’s study participants were given the opportunity to buy IKEA furniture they built themselves versus identical units built by others, they bid 67 percent more for their own creations. They did the same with Lego kits and paper origami. In all cases, people would pay more for things where their own sweat was one of the ingredients. This is the Maker’s Premium. It’s the ultimate antidote to commodification.

Take any niche and check out the new producers. Mountain bike parts, classic car accessories, cool vinyl “skins” for phones and other gadgets—they’re all seeing a wave of new micro-entrepreneurs sell­ing online. Although each market is different, what’s common about this new creative class is that they were once consumers who wanted something that didn’t exist before. So, rather than settle for what was on the market, they made something better themselves. And once they made one, it was increasingly easy to make more. And thus a small business emerged from the most passionate ranks of the con­sumer class.

What does “artisanal” mean in a digital world? In his 2011 book, The Alphabet and the Algorithm, Mario Carpo, an Italian architectural historian, argues that “variability is the mark of all things handmade.” So far, no surprise for anyone who has bought a tailored suit. But he continues,

“Now, to a greater extent than was conceivable at the time of man­ual technologies... the very same process of differentiation can be scripted, programmed, and to some extent designed. Variability can now become part of an automated design and production chain.”²²

Just consider the Web itself. Each of us sees a different Web. When we visit big Web retailers such as Amazon, the storefront is reorganized just for us, displaying what algorithms think we’ll most like. Even for pages where the content is the same, the ads are different, inserted by software that evaluates our past behavior and predicts our future actions. We don’t browse the Web, but rather search it, and not only are our search strings different, but different users get different results from the same search strings based on their personal history.

Writes Carpo, “This is, at the basis, the golden formula that has made Google a very rich company. Variability, which could be an obstacle in a traditional mechanical environment... has been turned into an asset in the new digital environment—indeed, into one of its most profitable assets.”

Information inside

But surely custom-made or bespoke suits and farmers’ markets have been around forever. What’s different now? The simple answer is that DIY culture has suddenly met Web culture. And the intersection of the two lies in digital design: physical products that are created first onscreen.

Walk into an Apple Store and look around you. All those shiny objects—all those beautifully designed and manufactured slabs of titanium, high-end plastics, and circuitry—started life on a screen somewhere. So, too, for a Nike store. Or a car dealership.

Physical products are increasingly just digital information put in physical form by robotic devices such as CNC mills and pick-and-place machines making printed circuit boards. That information is a de­sign, translated into instructions to automated production equipment. In a sense, hardware is mostly software these days, with products be­coming little more than intellectual property embodied in commod­ity materials, whether it’s the code that drives the off-the-shelf chips in gadgets or the Ç-D design files that drive manufacturing.

And the more products become information, the more they can be treated as information: collaboratively created by anyone, shared globally online, remixed and reimagined, given away for free or, if you choose, held secret. In short, the reason atoms are the new bits is that they can increasingly be made to act like bits.

The result is that we are now seeing what looks like what Joseph Flaherty, who writes the Replicator blog, calls a “Moore’s Law for Atoms.” The original Moore’s Law, named after Intel researcher Gor­don Moore, described the twenty-four-month doubling of processing power per dollar that has characterized the computer industry since the 1970s. That exponential growth comes from the phenomenon of “compound learning curves”: breakthrough discoveries in semicon­ductor research come frequently enough (about every three years) and build on their predecessors so effectively that progress accelerates at this breakneck pace.

Why do all industries not enjoy this pace of improvement? Be­cause semiconductors are still a relatively new field in the long arc of scientific research. They are built on the quantum mechanics and ma­terial science breakthroughs of early twentieth century, a remarkable period of discovery that opened an entirely new domain of physics. As Richard Feynman famously said, “there’s a lot of room at the bottom,” at the atomic level of matter, and we re Stilljust beginning to plumb it.

What is the analogy for manufacturing? Nothing so grand as a new physics. Instead, it is simply the combination of the technologies that the original Moore’s Law brought us: computers, digital infor­mation, the Internet and, most important of all, connected people.

Remixing the physical world

It’s easy to miss the magnitude of this shift. After all, from a distance the whole process of making things doesn’t seem so different. My grandfather designed his machines on paper and prototyped them by hand in his workshop. I design in CAD and send the files to be prototyped on my desktop fabricator or by robot machines in a remote service bureau. But at the end of the process, we still both have a pro­totype in hand. What’s the big deal in doing it my way?

The answer comes down to the unique qualities of digital informa­tion. It seems like such a small distinction: products shared as physi­cal things or products shared as digital descriptions of physical things. After all, if you’ve got to fabricate them to make them real, one way or another, who cares what form the instructions take?

But as we’ve learned over the past few decades, digital is different. Sure, digital files can be shared and copied Iimitlessly at virtually no cost and with no loss of quality. But what’s more important is that they can be modified just as easily. We live in a “remix” culture: ev­erything is inspired by something that came before, and creativity is shown as much in the reinterpretation of existing works as in original ones. That’s always been true (the Greeks argued that there were only seven basic plots, and all stories just changed the details of one or an­other of them), but it’s never been easier than it is now. Just as Apple encouraged music fans to “Rip. Mix. Burn,” Autodesk now preaches the gospel of “Rip, Mod. Fab” (Ç-D scan objects, modify them in a CAD program, and print them on a 3-D printer).

That ability to easily “remix” digital files is the engine that drives community. What it offers is an invitation to participate. You don’t need to invent something from scratch or have an original idea. In­stead, you can participate in a collaborative improvement of existing ideas or designs. The barrier to entry of participation is lower because it’s so easy to modify digital files rather than create them entirely yourself.

My grandfather was a lone inventor, not because he was especially solitary but because he had no mechanism for easy sharing. I may be no more extroverted than he was, but because my medium is digital, sharing comes naturally. When you share, community forms. And what community does best is remixing—exploring variation in what a product can be, and in the process improving it and propagating it far faster than any individual or single company could.

Think of a digital product design not as picture of what it should be, but instead as a mathematical equation of how to make it. That is not a metaphor—it’s actually the way CAD programs work. When you draw a Ç-D object on the screen, what the computer really does is write a series of geometrical equations that can instruct machines to reproduce the object at any size in any medium, be it pixels on a monitor or plastic in a printer. Increasingly, those equations don’t just describe the shape of a thing, but also its physical properties—what’s flexible and what’s stiff, what conducts electricity and what insulates heat, what’s smooth and what’s rough.

So everything is an algorithm now. And just as every Google search uses its algorithms to produce a different result for each person searching, so can algorithms customize products for their consumers.

For the 99 Teapots project, Architecht Greg Lynn designed one teapot in a CAD package, then let the software remix it to create ninety-eight others. Each was fabricated in a carbon mold, and tita­nium was exploded within it to create a unique teapot. (With a price tag of up to $50,000, they were more art piece than serving set, but the process was as interesting as the product.)

Lynn explained the point: this sort of variation in form is the es­sence of being a modern designer. In a 2005 speech at the TED con­ference, he explained the BMW design challenge. At any given time, the car company has scores of designs, ranging from the $30,000 300 Series to the $70,000 700 Series. All of the BMW cars should “look like BMWs,” which is to say there should be a family resemblance. But if the 700 Series is going to justify costing more than twice as much as a 300 Series, it can’t look too much like it. Instead, it has to look more like other 700 Series cars.

What factors determine “BWM-ness”? And what factors de­termine “700 Series-ness”? It can’t just be the mechanical specifi­cations—there must also be some ineffable aesthetic that is hard to describe in words but easy to see. Decades ago, the ability to do that defined a master designer, and perhaps if you work for BMW and Apple, companies defined by personal design vision, it still does. But for most companies today, it defines a master algorithm instead. Software is increasingly driving the design process, with the broad strokes created by the human eye but the details and variations all filled in by code following rules dictated by material properties and manufacturing efficiencies, easily remixed by others into any num­ber of variations.

Carpo explains what this represents: “Algorithms, software, hardware and digital manufacturing tools are the new standards of product design.... Unlike a mechanical imprint, which physi­cally stamps the same form onto objects, an algorithmic imprint lets outward and visible forms change and morph from one object to the next.”

Sound familiar? This echoes the “mass customization” promise of the first wave of Web retail, a decade ago. If a product is built on demand, why not have it designed on demand, too, or at least offer the consumer the ability to customize it according to taste? Dell’s success with bespoke computers a decade ago promised an era when every­thing from cars to clothes would be made and sold that way.

But it didn’t happen, at least not at the scale that everyone ex­pected. Cars, for example, are chosen primarily for their reliability. The more variability there is in the manufacturing process, the harder it is to keep the defect rate down. Without perfect Ç-D models of the customer (and telepathic understanding of their wearing preferences), clothes are hard to tailor predictably, which is why men still have their inseams measured in shops.

Today the canonical examples of mass customization are still a bit trivial, to say nothing of getting long in the tooth: Nike ID shoes (you can design a novel pattern on standard sneakers), custom-printed M&Ms, and the like. Having your name inscribed on the back of your iPad is hardly an industrial revolution.

And even Dell hardly does mass-customization anymore. Today you can only choose the standard models with their two or three choices of memory, CPU, hard drive, and video card options, and if you don’t pick the most popular combination (which Dell mass pro­duces, the good old-fashioned way), you’ll have to wait an extra two weeks for delivery. Car companies do the same. They all found that more variety meant more variability in quality and uncertainty in in­ventory. Given a choice between infinite options and products that are cheap, available, and reliable, consumers tended to go the safe, one-size-fits-all route.

Likewise, the examples where consumers are designing their own products online are rarely mass. Threadless (T-shirts), Lulu (self-published books), CafePress (coffee mugs and other trinkets), and others like them are thriving businesses, but they are platforms for creativity more than great examples of mass customization. They simply give consumers access to small-batch manufacturing on stan­dard platforms: shirts, mugs, and bound paper.

So I won’t be invoking “mass customization” much here. In­stead, what the new manufacturing model enables is a mass market for niche products. Think ten thousand units, not ten million (mass) or one (mass customization). Products no longer have to sell in big numbers to reach global markets and find their audience. That’s be­cause they don’t do it from the shelves of Wal-Mart. Instead, they use e-commerce, driven by an increasingly discriminating consumer who follows social media and word of mouth to buy specialty products online.

In a 2011 speech at Maker Faire, Neil Gershenfeld, the MIT pro­fessor whose book∕v√Z>.∙ The ComingRevolution on TourTesktop antici­pated much of the Maker Movement nearly a decade ago, described his epiphany like this:

I realized that the killer app for digital fabrication is personal fabrication. Not to make what you can buy in Wal-Mart, but to make what you cant buy at Wal-Mart.

This is just like the shift from mainframes to personal com­puters. They weren’t used for the same thing—personal com­puters are not there for inventory and payroll. Insteadersonal computers were used for personal things, from email to video games. The same will be true for personal fabrication.²³

Small batches

Blogger Jason Kottke wrestled with what to call this new class of entrepreneur ship, these cottage industries with global reach targeting niche markets of distributed demand. “Boutique” is too pretentious, and “indie” not quite right. He observed that others had suggested “craftsman, artisan, bespoke, cloudless, studio, atelier, long tail, agile, bonsai company, mom and pop, small scale, specialty, anatomic, big heart, GTD business, dojo, haus, temple, coterie, and disco business.” But none seemed to capture the movement.

So he proposed “small batch,” a term most often applied to bour­bon. In the spirits world, this implies handcrafted care. But it can broadly refer to businesses focused more on the quality of their prod­ucts than on the size of the market. They’d rather do something they were passionate about than go mass. And these days, when anyone can get access to manufacturing and distribution, that is actually a viable choice. Wal-Mart, and all the compromise that comes with it, is no longer the only path to success.

The collective potential of a million garage tinkerers is about to be unleashed on the global markets, as ideas go straight into production, no financing or tooling required. “Three guys with laptops” used to describe a Web startup. Now it describes a hardware company, too. “Hardware is becoming much more like software,” as the MIT pro­fessor Eric von Hippel puts it.

The Web was just the proof of concept of what an open, bottom-up, collaborative industrial model could look like. Now the revolution hits the real world.