A new wave of entrepreneurs is proving you don’t have to go overseas to churn out a product – or a profit. Thanks to flexible digital tooling, shared R&D spaces and crowd-sourced funding, it’s easier than ever to turn a good little idea into the next Big Thing. By Steven Leckart | Photographs by Joe Pugliese
‘This!’ shouts Patrick Buckley, ‘is the Big! Robot! Room!’ Buckley is whisking me through the 930 m² factory his company, DODOcase, operates in an industrial area of San Francisco. We’ve stopped to admire the 32-year-old CEO’s “Ferrari”. It’s not an Italian sportscar, but a very loud programmable CNC router with about the same footprint, and half the price tag.
Right now, the R870 000, red-and-white SCM Pratix is cutting precision details into 18 bamboo frames held in place by a vacuum-sealed jig. Ultimately, each 20 x 25 cm rectangle will be glued into a handmade book cover. The final product: a R530 iPad case.
Resembling a Moleskine notebook, the DODOcase has exploded in popularity since debuting alongside the iPad in April 2010. Within a month, orders spiked from 10 to 900 a day. Retailers like J Crew carry them, and President Obama keeps one on his desk.
DODOcase hasn’t always had a big robot room. Or its own bookbindery. Or 25 full-time employees. When Buckley and co-founders Craig Dalton and Mark Manning started the company, it seemed more like a hobby than an assembly line. They cut bamboo on routers at the DIY hackerspace TechShop, outsourced covers to a local bookbinder, and assembled the cases in Buckley’s basement. Three years later, DODOcase has grown into a model of success for a new breed of small-scale manufacturers.
They’re not alone. Manufacturing used to require lots of capital and scale. But new technology is making it easier, faster and cheaper to transform an idea into a business. But new technology is making it easier, faster and cheaper to transform an idea into a business. Free, easy-to use computer-aided design (CAD) programs and inexpensive 3D printers help entrepreneurs prototype quickly. Shared R&D spaces such as NextFab and TechShop give anyone access to expensive tools for around R900 a month. And crowd-funding Web sites such as Kickstarter let innovators raise capital quickly while maintaining control of their companies.
Starting a business still takes plenty of sweat equity and a bit of luck. But if you want to graduate from spare-time tinkerer to full-time manufacturer, there has never been a better time to do it. Here are five essential strategies for getting started.
Lesson 1: Build connections
Manufacturing hasn’t always been a welcoming industry for small-batch entrepreneurs. Equipment and resources were most available to companies that produce hundreds of thousands of products, leaving the little guy on his own. The rise of shared workspaces for innovators is changing that.
In 1997, Edmund Villarreal began crafting magnesium fire starters on his apartment balcony in Euless, Texas. Featuring a handle cut from specialty woods such as hickory, dogwood and mesquite, the screwdriver-size All-Weather Firestarters weren’t that complicated to produce. He had his own magnesium cutter and band saw. But it was time-consuming. At best, Villarreal could crank out one Firestarter every 20 to 30 minutes. After a decade of doing odd jobs during the day and making Firestarters at night, he had managed to sell a respectable 20 000 units, but he was maxed out.
“I needed to manufacture these things on a bigger scale, but I just didn’t know how,” he says. In 2009, after moving to Raleigh, North Carolina, Villarreal posted an ad on Craigslist, hoping to find an out-of-work engineer to cut his magnesium. He scored one – and more. The engineer turned out to be a member of TechShop.
Villarreal joined TechShop, too, and other members soon offered helpful tips: don’t swap out bits every time you drill a countersinker. Try a bit with an integrated countersink! Don’t polish and cut the magnesium one piece at a time. Automate!
Villarreal gained instant access to high-end equipment and skilled tradespeople. “I’m learning how to work smarter, not harder,” he says. Today, whenever he gets an order, he hires an ad hoc team of four part-time engineers to set up and run a custom line at TechShop. The guy who programs the ShopBot PRSalpha router gets R450 an hour, but it’s worth the money. The team produces one Firestarter every 10 minutes. Sales have picked up, too. In 2012, Villarreal sold 5 000 Firestarters.
Lesson 2: Keep inventory lean.
Matching supply to demand is the most critical balancing act in business. Ramping up production requires investments in equipment, materials and staff. For companies that are just getting off the ground, gauging demand often amounts to a lot of guesswork.
The fundraising Web site Kickstarter fundamentally flips that equation, with small investors promising money up front for early access to goods that haven’t been produced yet. It’s a transformative idea, and it’s launched a lot of little companies into the big time faster than they expected. Last year, the smart-watch start-up Pebble began a campaign on the site to raise R900 000 but ended up raising 100 times that amount. The company suddenly had an obligation to produce 85 000 smart watches. So instead of continuing in its small factory in San Jose, California, where the team had made its batch of 1 500 samples, the founders turned to Dragon Innovation, a firm that helps companies outsource manufacturing to Asia.
But an initial flood of orders doesn’t guarantee long-term success. In 2011, Dave Petrillo and his partner started a Kickstarter campaign for Coffee Joulies – stainless-steel beans filled with a phase-change material to keep coffee at a desired temperature. They planned to raise R850 000 to cover half the initial cost of tooling, but the company ended up with R2,7 million. That’s when they decided to set up manufacturing in a 110-year-old silverware plant in Sherrill, New York. Six months later, the company had filled the 4 818 Kickstarter orders. But then it hit a speed bump.
After the holidays, demand slowed but the factory didn’t, and soon the company had a backlog of 32 000 unsold units. “Just when we got into a comfortable work pace, we had to stop. We underestimated the seasonality of our product,” Petrillo says. Good PR saved the day: after a January 2013 appearance on the ABC show Shark Tank, sales boomed, and now Joulies are on back order.
The perfect balance, according to DODOcase’s Buckley, is to link production to demand. “We control the manufacturing process. We can turn our machines on and off whenever we want,” he says. “We won’t ever have more than two weeks of inventory.”
Lesson 3: Get parts-savvy.
Even the leanest production process relies upon a supply chain of parts and materials. Before the Internet, finding reliable suppliers could take years of relationship building. Now it takes seconds. Web sites such as MFG.com automate the search for manufacturers and suppliers, and build in quality control with user reviews and a five-star rating system.
The supplier culture is changing, too. Entrenched vendors are more willing to work with start-ups. The runs may be small, and there’s no guarantee that the products will lead to further orders, but vendors often have no choice. Many big manufacturers have cut back since the economic downturn and factories are seeking to make up the loss of big orders with many smaller projects. The result is the growth of an ecosystem that lives and breathes small batches.
“Finding vendors for this project has been easier than anything I’ve ever worked on,” says Benjamin Cohen, who founded a bicycle-headlight company called Blink/Steady in Brooklyn, New York, with two robotics engineers who formerly had freelanced for Nasa. “The old guard is stepping back and the younger people at these companies are willing to work with you.”
In July, Blink/Steady filled the first orders of its R850 (now R1 100) bicycle headlight, a multi-sensor device that automatically turns on when the bicycle is in motion or it’s dark outside. The product is constructed with a sleek aluminium body (anodised in Pittsburgh) that’s been laser-etched (in Wallingford, Connecticutt), outfitted with a waterproof O-ring (fabricated in Lancaster, New York) and stocked with a circuit board (assembled in San Jose and programmed in Brooklyn).
Knowing how your parts are designed is just as important as knowing where to get them – especially when it comes to electronics. George Yu is the CEO of Variable Technologies, a start-up in Chattanooga, Tennessee, that makes a wireless iPhone accessory called NODE. “The whole world of circuit-board assembly is a little mysterious to small-batch creators,” he says. “Vendors are not geared towards amateurs or beginners at all.”
NODE, which packs together sensors for measuring data such as light, motion and temperature, is only 7,5 cm long, so the circuit board had to be designed tightly. Yu had earned a PhD in electrical engineering from Georgia Tech, but he had no clue how circuit boards are massproduced. So he spent days watching videos of circuit-board pick-and-place machines on YouTube and scouring Web sites for written information. He arrived at two ahas: 1) pick-and-place machines employ different-sized nozzles to pick up parts, and 2) changing those nozzles impacts cost.
Therefore, Yu realised, if he designed every part and resistor on his circuit board to be roughly the same size, there would be no need to change nozzles. Ultimately, that insight streamlined the process of manufacturing NODE, saving Yu time, effort and money.
Lesson 4: If America can build it, build it in America.
By the same token, you might find that it’s both easier and cheaper to do it right here in South Africa. Your heart says let’s keep it local, while the economics of international labour markets say build in China. But as Americans Dave Petrillo and David Jackson discovered when they decided to make Coffee Joulies domestically, the logistics of small-batch manufacturing are more complicated than that.
“Our decision wasn’t patriotic, as nice as that would sound,” Petrillo says. “But manufacturing in the US can be an asset.” For starters, there’s no language barrier. No 12- to 15-hour time change to contend with. And no international airline tickets to eat into your budget. Whether you’re in Asia or Gauteng, setting up a new manufacturing line requires a lot of trial and error. Then comes more back-and-forth once production begins. Long distances and time-zone changes can clog up the process.
Scott Miller, the CEO of Dragon Innovation, is an expert in offshore outsourcing, but even he tells many small start-ups not to produce in Asia. He points out that many factories in China impose a “magic number” threshold that small companies can’t meet. Volume producers may require a minimum order quantity (MOQ) of 5 000. Even if your small batch meets that number, do the maths to make sure you’re really coming out ahead by outsourcing abroad: Miller assembles a decision matrix weighing factors such as margins, labour rates and the price of components.
Products that have to ramp up quickly, such as Pebble, may need to take advantage of the massive manufacturing capabilities in Asia. Likewise, products that require a lot of manual labour are cheaper to make in China – but that may not be true for long. In 2001, the average manufacturing wage there was R5 an hour. Today it’s closer to R27 an hour. The Boston Consulting Group predicts that there may be no economic advantage to producing in China by 2015. That’s why even big companies are following small manufacturers back to North America. Companies such as Ford and GE are “reshoring” some of their production to the US.
For some small manufacturers, the best strategy takes advantage of the US’s friendly trade relationships with its neighbours. 3D Robotics, launched in 2009 by Chris Anderson, then Wired editor-in-chief, sells both assembled and DIY unmanned aerial vehicles. The company initially used circuit boards assembled in China. But by 2010, Anderson’s cofounder, Jordi Muñoz, had learned how to construct custom boards at home. After months of soldering by hand, Muñoz spent R27 000 on eBay for a used pick-and-place machine to create the boards faster. The company graduated from a tiny garage to a warehouse in San Diego.
When it needed to expand to a second facility, the founders looked south – to a 1 300 m² factory in Tijuana, Mexico, where Muñoz was born. In 2012, Anderson quit his day job. “The day I finally decided maybe it was time to consider leaving Wired was the day Jordi sent me a picture of our forklift,” Anderson says with a laugh. “Real companieshave forklifts.”
Lesson 5: Prepare to pivot.
Even if you achieve success with your first small-batch widget, a number of factors can throw a kink in your business plan. Unexpected competitors, market changes or technological shifts can demand new products or a change of design. Anderson says that the same general-purpose, digitally controlled manufacturing tools that enable small-batch production also allow entrepreneurs to be nimble. “With older, analogue methods such as injection moulding, you get great economies of scale,” he says. “But the downside is that once you’ve made a part, you don’t dare change the design – because it’s expensive. With digital manufacturing tools, you can change the design every day.”
Buckley from DODOcase agrees: “I come from the start-up world where we’re constantly thinking, ‘What’s the pivot?’ If all of a sudden our iPad business dried up, we would start trying new stuff.” He’s already doing just that. As my tour wraps up, I notice a surfboard sitting in a dusty corner of DODOcase’s workshop. Buckley tells me he assembled the 1,7 m board using the same bamboo he uses for iPad cases. Eventually, he wants to commercialise a line of DIY, home-assembly surfboards.
All he has to do is upload a new file to his Ferrari. Instead of automatically cutting 18 DODOcases, that SCM Pratix could just as easily mill the dimensions of, say, eight 1,8 m surfboards. “That’s my dream,” he says. That could be the ultimate attraction of the flexible, small-batch model: Once you succeed with your first product, it’s even easier to get cranking on your next great idea.
Launch a start-up in 7 easy steps
With the right tools and technology, a smart guy like you can design, prototype and manufacture the product of your dreams. Just follow this plan for success!
Step 1: Design
The first step is to perfect your world-changing widget. Computer-aided design (CAD) software used to be pricey and hard to use, but new tools such as Autodesk 123D Design, Blender, Google SketchUp and Tinkercad make designing virtual 3D objects easy. Or build your widget by hand, and use Autodesk’s 123D Catch to convert digital images of it into a CAD file.
Step 2: Prototyping
Good job! You’ve figured out the basic shape of your widget. Now it’s time to build a prototype to test and show off to investors. You can do it the old-fashioned way – in your basement with box cutters and glue – or you can take advantage of a wealth of new resources and tools. Desktop 3D printers, laser cutters and CNC routers can whip out precision parts – and they have become cheap enough for ordinary citizens to own. If you’d rather not spend your seed money on equipment, you can send your CAD files to a 3D printing service or join a maker collective in your area.
Step 3: Financing
To turn your prototype into a product for the masses, you need a healthy injection of start-up cash. That used to mean handing over a stake in your company to venture capitalists, but the crowdsourcing site Kickstarter lets you fund up while maintaining full ownership. You can lure potential investors with a sliding scale of incentives, promising backers anything from karmic satisfaction to discounts on your first batch of products. Alternatively, talk to a friendly venture capitalist or even the appropriate government department: visit the Institute of Inventors and Innovators Web site for advice: www.iii.org.za
Step 4: Sourcing
Online ordering has democratised the sourcing of parts and materials. Vendors will sell you stainless steel, aluminium and brass by the bar or by the truckload. Thanks to the ubiquity of the smartphone, electronics parts such as accelerometers and GPS antennas are now affordable. Buy them piecemeal or in volume from suppliers via the Web.
Step 5: Staffing
Even with today’s high unemployment, finding skilled workers is a challenge. Try starting your business at a hacker space with like-minded tinkerers to help you fill key positions. Big staffing companies can help fill temporary positions. Or you can farm out tasks to an elastic army of freelancers.
Step 6: Manufacturing
People love your widget! Now that orders have skyrocketed, demand can quickly outpace the production capacity of your living-room assembly line. That’s a good problem to have, but ramp up in a hurry or you risk alienating customers. If you’re looking to set up your own factory, you can search for accommodation in an industrial park that meets your requirements (think about access, travel distances, availability of essential services, security and all the other factors that could affect your business). Or you can seek out a contract manufacturer anywhere in the world using www.mfg.com. The site can help you find an electronics plant in Mexico or a metal fabrication facility in Iowa.
Step 7: Sales and marketing
Getting your product on to Makro’s shelves might seem like a victory, but big retailers demand high volume and deliver low margins, so many small manufacturers sell directly to consumers. Online services take care of many of the details (hosting, payment processing, security) in exchange for a cut of each sale or a monthly fee. Next, get the word out through your social network. Talk up your product on Facebook and Twitter.
How to talk factory:
Now available for the cost of a nice bicycle, 3D printers can turn almost any CAD creation into a physical object. Sophisticated desktop units such as this Maker- Bot Replicator 2X use dual extruders to print in two colours of plastic. (See our recent CES report-back for an example of what’s coming.)
3D printers are extremely useful but for fashioning quick and precise parts from materials such as acrylic, wood and metal, nothing beats a laser cutter. The machines are expensive but they are available at shared R&D work spaces. If you can’t find one, why not set up one yourself?
CNC routing/ milling
These two machines are similar in concept but they operate differently. Routers direct a rotating bit to shape wood and plastic, while mills usually move the material (mills are good at cutting metal). Both are the inverse of a 3D printer, subtracting material instead of adding it.
Pressing and stamping
A standby of large-scale industrial manufacturing, machine presses force the manufacturing material against a pre-made form. Stamping machines cut out shapes much like a cookie cutter does. Both kinds of machine can produce parts much faster than printers and milling machines, but they are not as flexible
Vacuum forming is great for making plastic parts such as electronics casings and packaging, which are heated, vacuum-sealed to a mould, then cooled. Hydroforming, which uses liquid pressure instead of a vacuum, is better for metals and other heavy materials.
Rather than pressing material against a mould, an injection moulder forces material into one. These high-volume machines make parts out of glass, metal or plastic.
Online retailing may be frictionless, but sometimes you’re out in the physical world. Technologies such as Square turn a smartphone into a credit-card reader. We may not be quite ready for it in South Africa, but it will come.