Trending: Elizabeth Warren takes aim at Amazon’s taxes again during speech at Seattle rally
Bigstock Photo

[Editor’s Note: Anup Chathoth is CEO of Seattle startup Ubi Interactive, maker of virtual interactive whiteboard technology and a graduate of the Kinect Techstars Seattle accelerator. In this guest post, he shares lessons learned in the hardware business.]

If you talk to entrepreneurs or investors, or anyone else in the startup ecosystem, the most common refrain when it comes to hardware is that it is hard. There is some truth to this, given the inherent challenges of the hardware business.

  • Hardware requires significant capital expenditure to produce a working prototype. A laptop and a couple of hackers cannot make it happen. You need to build actual physical products, which cost time, money and require an ecosystem of partners.
  • Product iteration cycles for hardware consume significantly more resources and time than software products require.

This is not to say that things haven’t changed in recent years. Low-cost development platforms such as Raspberry Pi enable quick prototyping. Standardization in areas such as wireless radio modules eliminate the need to develop custom modules. Crowdfunding provides an excellent way to validate early market demand and also raise money for inventory.

But even with these changes, hardware is still a brutal business for many startups. A study from CB Insights analyzed 400 failed hardware startups, and found that the top two reasons they fail are 1) lack of consumer demand and 2) a high burn rate.

In other words, many hardware startups raise money on an idea, not on a product. The actual product sometimes ends up too costly (Coolest Cooler) or even impossible to manufacture (Lily Drones). Even if startups end up shipping the product, it often disappointed customers, hurting demand. Case in point: headphone maker Human, the Seattle startup that failed after five years of development and a reasonably successful crowdfunding campaign. Helmet maker Vicis was another promising Seattle hardware startup that also ran out of money.

Coolest Cooler
(Coolest Cooler Photo)

Software and SaaS startups avoid these early, expensive mistakes by applying the lean startup method, shortening product development cycles and iterating rapidly based on feedback to achieve product-market fit. This approach is even more important in hardware, because every mistake costs more time and money.

Hardware startups get tons of tips for being lean. But the problem is that this advice is not actionable. A duct-taped Raspberry Pi prototype is not enough as a minimum viable product (MVP). You need to develop a prototype that appears almost production ready, is stable enough to garner reliable feedback, and demonstrates the core value of your product. You need to get this into the hands of your customer and start iterating your product before you burn through your capital.

This is possible. Here is the playbook every hardware entrepreneur, especially in the consumer hardware and IoT space, needs to follow.

Tweak, don’t make: When the original team at Tesla set out to make Roadster, they very soon realized that they did not have the knowledge or resources to build an entire car. Instead, they’d simply build on top of, and within, an existing car, the Lotus Elise. This holds true for most hardware startups. Do not set out to build an entire car. Build only your engine, that key new feature that delivers your core value.

Given how standardized hardware has become these days, you will find a product in the market that matches very closely what you want in your MVP. All you need is to fit your own engine inside.

Go to China: This may be the most controversial part. Don’t waste your time trying to figure this out yourself, sitting in Seattle or Silicon Valley. Once you know your MVP and the product you want to tweak, take that flight to Shenzhen. The city is packed with independent design houses (IDH), each specializing in different verticals. They, of course, have engineering resources for software and hardware. But most importantly, each of IDH has close relationships with manufacturers who are well versed in the supply chain ecosystem. Everything they produce is designed to manufacture. They have figured out 90% of the details of the product you want to build.

Ideally you should find the IDH that designed the original product that you are tweaking or appropriating. Be prepared to spend weeks with them. Be prepared to sleep in their office, be prepared to drink with them every night. Learn everything you can about the design and components of the product you are building.

Design to manufacture the prototype: Come home with a ready-to-manufacture version of your MVP. It may still have 3D printed internals and hand-soldered components. But it should be already designed to manufacture and ready to be “play tested” by your customers.

Keep iterating: Once you do play testing or field-testing, go back to China again. Rinse and repeat until you know your MVP is ready to go to mass production (MP). You typically will have to go through multiple phases of Product Validation Testing (PVT) before you are ready for MP. Since you are tweaking a product that went through MP before, you will be surprised by how much you save in time and money during each of these iterations.

You now have given yourself the advantage only software companies had before. You are in a position to iterate your MVP quickly and cheaply, incorporating customer feedback while at the same time keeping it ready to manufacture.

So hardware is not hard. It’s different, and doable, if you follow the right playbook.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Comments

Job Listings on GeekWork

Executive AssistantRad Power Bikes
CTO-in-Residence // Co-FounderALLEN INSTITUTE FOR ARTIFICIAL INTELLIGENCE (AI2)
Technical Co-Founder // CTO (Stealth Startup)ALLEN INSTITUTE FOR ARTIFICIAL INTELLIGENCE (AI2)
Find more jobs on GeekWork. Employers, post a job here.