Gadget fans can't wait to get their hands on Myo, an armband that controls PowerPoint or Keynote presentations, videogames, toy drones, music and various software programs through simple hand gestures. To outsiders, its creator Thalmic Labs seems like the textbook definition of a successful startup: Y Combinator alum, 30,000 pre-orders secured and $14.5 million raised from investors – all before its Myo officially launches this September.

But during an exclusive tour of Thalmic Labs' research and development lab, the company's CEO, Stephen Lake, told Entrepreneur how his team took very deliberate steps not to accept any actual money from consumers. He's also had no qualms tinkering, tweaking and trying to perfect Myo – even as its delivery to thousands of developers has been delayed by months. "We could have shipped these," says Lake while holding up an older model of Myo, "but I guess we're all perfectionists and we weren't fully satisfied."

Unusual? Certainly. But there's good reason. Here's what your venture could learn from the team behind Myo, which might just become that next big thing:

Picture the bigger picture. Ask some individuals what inspired their invention and they'll start by telling you about a specific problem they wanted to address – not a category of products they aimed to sell. For Lake and his two co-founders, their research into the next evolution of computing found that wearable devices was set to explode. "But one of the big challenges that is unsolved is how do you interact with this next generation of computers all around you?" says Lake.

"People always thought in the future you're going to have these kinds of interfaces but no one ever really thought about how you're going to interact with them." He notes that voice- and camera-based gesture controls don't really work well for the types of applications his team has in mind, including the use of smart glasses, heads-up displays and augmented reality (AR) eyewear. And his team is taking a diversified approach by pursuing both consumer as well as business applications, having recently announced on its website that it is working with leading partners to deploy wearable computing solutions in sectors such as healthcare, construction and outdoor work. "We wanted a more subtle interface that would allow you to interact with those types of devices. That's where we came from."

Leverage unique areas of expertise. Before they got started in 2012, the trio who launched Thalmic Labs studied at university together for several years. Matthew Bailey had studied pattern recognition, which has helped him develop the underlying machine learning behind Myo, while Aaron Grant focused on software engineering, an obvious asset today.

Lake, meanwhile, studied hardware, biomechanics and electromyography tests that analyzed the muscle activation patterns of athletes. "We made the connection: If we can detect these muscle activation signals, couldn't we then use them to interpret the motions you're making in gestures?" says Lake. "That was the basis for how Myo works today then we spent six months building out early prototypes to prove that was actually possible."

Go against conventional wisdom. To secure its 30,000 pre-orders, for which consumers have been paying $149 once they've actually received the product, Thalmic Labs created a campaign on its own website – and deliberately avoided launching one on crowdfunding sites like Kickstarter or Indiegogo. "One of the reasons we chose not to take money up front was so that we had some flexibility in saying the bar for us in shipping these is [based on] quality and our own standards," says Lake. "Pre-orders can be an early, positive way to go because it gives you more flexibility – you're not taking people's money hostage if you're not sure when it will ship [or getting] negative PR that way."

Consider an approach. Then reconsider it. Originally, Thalmic Labs considered 20 to 25 hand gestures but discovered in early tests that people didn't want to learn that many. (Today there are five core gestures, including moving the hand side to side and opening or closing a fist.) The company also planned to ship thousands of alpha units this past December to developers but then selected just 1,000 applicants, who have become a small feedback group to help Thalmic's programmers improve their algorithms.

At the same time, employees have been upgrading Myo's design. The issues? For starters, it was too thick: "We certainly wanted something that was the thickness of an iPhone, or thinner, so you could fit it under your sleeve so it's not obtrusive. This," Lake adds, holding up a chunkier model that was about twice as thick, "is just not the aesthetics we wanted to deliver in the end."

The old Myo also featured separate plastic parts that Lake worried could easily break, while the latest is molded in one piece, fits a wider array of arm sizes and includes a more rugged design. It came at a cost, however: a different manufacturing process. (Those who didn't pre-order Myo for $149 will now have to pay $199.)

Plan the next move. At the end of the day Myo is just a single device. While Lake won't discuss whether he has plans to develop a broader suite of gadgets, he says he's focused on making sure Myo becomes compatible with different programs beyond those that are easiest for consumers to visualize at launch. Hence the push into using it as a touch-free, voice-free way to control enterprise applications, including those on Google Glass, Recon Jet and Epson Moverio.

"One of the challenges we have is how do you explain to the consumer what this is? You don't walk into a Best Buy saying, 'I want a gesture-controlled armband,' so we've tried to find a good comparison: When the iPad was first launched, Apple was good at showing you specific things it could be used for – even if they're not the things that, long-term, you might be most interested in. But you understood how it worked."