This installment is a report on a highly unusual organization at which this innovation and stimulation are underway. It’s unusual because it represents the collaborative efforts of a very old, very large, very successful worldwide corporation, and a community of very new, very small, very local entrepreneurs.
This is of course the FirstBuild “microfactory,” established in Louisville, Kentucky, two years ago as an offshoot of GE’s Appliances division, whose large “Appliance Park” manufacturing center is on the southeastern edge of Louisville. To give an idea of its micro-ness: GE Appliances as a whole has something like 12,000 employees; the full-time staff of FirstBuild is about 20.
What was the big idea here? From GE’s point of view, it started with an effort to reduce the “mind-to-market” cycle time, as FirstBuild’s CEO, a longtime GE veteran named Venkat Venkatakrishnan, told me when I visited. He explained that different products have naturally different design-and-manufacturing cycle times, due to the volatility of their respective markets and how hard and costly it is to retool them. (I spoke with other people at FirstBuild, but Venkat, as he is known, is the only one with whom I spoke for on-the-record quotation.)
On one extreme for fast cycle-time is clothing, with styles going in and out of fashion in a matter of weeks or months. Not far behind are smartphones. On the other extreme are commercial aircraft, which may first enter service decades after the start of the design process, and which involve tens of billions of dollars in capital-cost commitment.
Appliances like those GE has made are in between. Refrigerators, freezers, dishwashers, clothes washers and dryers, air conditioners, and related heavy-duty, long-life-span products typically take about four years “mind to market”—from the first design concept or wish-list to the first on-sale item in the store. But no matter the cycle’s duration, weeks or decades, faster would always be better.
Faster means an edge on the competition. It means less tied-up capital. It means a better ability to offer new features or efficiencies. It means quicker response to shifting tastes and market trends. Years ago I came across and wrote about the retired Air Force colonel John Boyd and his concept of “the OODA Loop,” for Observe, Orient, Decide, and Act. Boyd based on the concept on his experience as a fighter pilot, where he found that the edge didn’t go to the pilot with the fastest plane, but rather with the one who could respond most quickly to an opponent’s move. The principle ripples through all forms of military competition, sporting contests, political campaigns, verbal sparring and debates, and of course business competition as well.
“Jeff’s challenge to us [this is Jeffrey Immelt, the CEO of GE] was, make it shorter,” Venkat told me at FirstBuild. “If it takes four years—for the feasibility studies, and the budget cycles, and the design and production—by the time you get to market, the technologies have changed, the market has shifted. So Jeff asked us what we could do to shrink the time, to get us on the path from years to months.”
The GE team began looking at startups, which of course must be nimble to survive. And they were re-impressed with the reality that size itself was an impediment to faster cycle time. Not necessarily the size of the corporation, whose engineering and financial resources could in principle permit fast reaction; rather, it was the size of their necessary production runs. If your system is set up to deliver tens of thousands of units to tens of millions of customers, it will automatically be biased toward caution. Any bet is a very expensive bet—in production outlays, in staff commitment, in opportunity cost. If the bet goes bad and the product flops, the costs will be paid for years, and an even-more-cautious approach will color subsequent decisions. The bigger the stakes in any product investment, the more deliberate the selection-and-design process is likely to be, and the narrower the range of possibilities it is likely to consider.
The low-volume idea was to make manufacturing faster and more varied, by dramatically lowering the output level at which a product would pay off. You can think of the analogy in TV programming. When there were only three networks, the only shows that could make it were those that could draw a mass audience. But with cable, and then with online videos and podcasting, much smaller and more specialized audiences were “large enough,” and we went from mass-cult Gidget and Get Smart to the much more varied, niche-sensitive, and more interesting modern range from Breaking Bad to reality schlock. Another obvious comparison would be the shift from the standardized, bland-ized mega-brewery U.S. beer world of the mid-20th century, to the rapidly diversifying craft-brew renaissance of today.
Venkat used a writing world analogy. “Suppose you were a publisher with only four authors,” he said, because you could only publish books that would be mega-sellers around the world. “How interesting would that be, and how long would you stay in business?” That is sort of the situation of large-volume manufacturers. “We’re more like the publisher who sees anyone walking down the street as a potential author. We’re a low-volume, high-variety publishing firm.”
What does this mean, in practice? Without going into all the details, essentially the approach is a combination of the well-known, intangible digital tools of the modern era, with some very tangible, less-known-by-the-public maker-era tools that (as I argued before) are transforming production.
That is: the products coming out of this microfactory use cloud-centric digital techniques we’re all aware of including crowdsourcing, online collaboration, crowd-funding, online sales, open-source coding and design. But they also use the new production techniques, from the real world of real hardware, that have become available only in the past few years—and that keep improving thanks to Moore’s Law. These include 3D printers, laser cutters, low-cost but high-sophistication multi-axis machine tools, and a range of other devices. Together these tools allow people in smaller, less formal, much lower-cost workspaces to design, test, refine, and manufacture items that previously would have come required factory production lines—and to find audiences that collaborate in the process of design and, so far, have provided eager markets.
“We are trying to overcome the selection bias of needing to scale up for big-volume production,” Venkat told me. “We don’t have to design to a spec. We are making in small-batches, low volumes. We can make one, then the next one, then the next one. We want to be open to as many ideas as we can.”
I made it a little too obvious that I was most interested in FirstBuild as a concept—what it showed about the maker era, what it might mean for locally based entrepreneurs. Venkat made clear that he cared about those things too—but that from his and GE Appliances’ perspective, this was an entirely serious business proposition, as a way to explore new markets and bring new products to them. If a new item sells in the tens or hundreds, that’s fine in itself. If it seems destined for a volume of more than a thousand units, “we can scale it up to the mother ship, to Appliance Park,” Venkat said. Meanwhile the items come out in units of tens or hundreds. “We are like the pirates, and Appliance Park is the navy,” Venkat said. “We’re the agile, exploring part of the empire. If we find something valuable, we call in the navy. Meanwhile, we explore.”
What do they make? “If it’s smaller than a shoebox, you buy it from China,” Venkat said, with an oversimplified but handy rule of thumb. “Bigger than a shoebox, you make it here.” You can read about a range of products they’ve put online—crowd-sourced in design, crowd-funded in many cases—at their website here. Most of them I saw were culinary-themed: a new kind of “nugget” ice maker, a portable smoker, a sous-vide cooker, new refrigerators, a compact pizza-oven that is hot enough to produce very good pizza, and, most startling to me, a high-speed home cold-brew coffee maker, which can produce (very good) cold-brew coffee within 10 minutes, rather than overnight. I won’t attempt to explain it but will say: When it goes on sale, I will buy one as a present for one of my cold-brew-loving offspring.
Will FirstBuild prosper under Haier? I don’t know. Will enough of its current products take off to sustain a new manufacturing wave? I can’t know that either.
But when people say, “Americans don’t make things any more,” you should know: Actually, they do. In Louisville, as you see—and in a range of other places we’ll discuss in the next installment.