The Lean Hardware Startup: From Prototype To Production

Editor’s note: Cyril Ebersweiler is the founder of the pioneering hardware startup accelerator HAXLR8R (which is now looking for applicants) and Partner at SOSVenturesBenjamin Joffe is an expert on startup ecosystems, angel investor and Advisor at HAXLR8R. Both invest in companies around the world and spent over a decade in China and Japan. This is Part 1 of a series. 

If the printing press was about “anyone can read,” the web about “anyone can write,” the hardware ecosystem changed enough to say today “anyone can build.” This idea – that anyone can build – is the cornerstone of the new “lean hardware startup.”

Yet, despite successes like SquareJawbone, and Fitbit, hardware startups continue to look daunting to entrepreneurs and investors alike. As investors in over 30 of them through our hardware-focused accelerator HAXLR8R (and in a number of startups outside HAXLR8R, too), we would like to share some ideas on how the landscape has changed for hardware entrepreneurs, and how it is now possible to be “lean” in hardware, too.
leanhw5

Lean hardware

The first challenge for hardware entrepreneurs is to get from your first prototype with 3D-printed parts, duct tape and cardboard to production-ready.

Steve Blank, a key inspiration in the Lean Startup movement, famously said “No business plan survives contact with the customer.” If you’re doing hardware the lean way, “No hardware plan survives contact with a factory” should be your mantra.

Design with the right components

Hardware often starts with a “bag of parts.” This won’t cut its time to hit the factory. Non-standard components make it nearly impossible to manufacture on a larger scale. If you don’t want to be doing all the assembly work by hand with your friends in the style of Steve-Jobs-the-movie, you’d better look for the right parts as soon as possible.

A prototype is ready when it can be manufactured
leanhw6
You got the right components? Great. But you’re not quite done yet: if your prototype can’t be made or even assembled, you’re toast. Luckily, factories often know better what can or cannot be done and can help you figure it out.

Of course, it works much better if you can be on site frequently to discuss it with the people actually doing the work rather than send long emails with specifications. You can thus iterate on the design of your prototype much faster.

Many of our startups saw the design of their products evolve, the sous-vide machine Nomiku is probably one whose changes were the most radical between the moment they joined HAXLR8R and the moment they went into production (read their story here on TC).
leanhw8
Manufacturing ability is one thing but costs are often a determining factor. A quote from Alibaba is not a reliable estimate nor is asking for “Apple quality” for half the price of a reasonable request to a factory. Avoid being overcharged or being laughed at by getting a better sense of your bill of materials and the manufacturing process.

Your factory is your most important partner

Okay so you have the right components and you know your product can be made and assembled. Are you going to select the factory and handle the relationship yourself?

It’s easy to make excuses for this one: it’s complicated, it’s in China (or Mexico, or elsewhere), you don’t speak the language, it takes time.… And where to start? Overall, it is scary.
leanhw9
As a result, it is tempting to bring in a third party to handle the relationship with factories. Beyond reducing your margins, the problem is that those companies essentially end up doing both audit and consulting. You know the result of this.

You can’t do hardware for long not knowing anything about manufacturing. So while you might need advice to get started, there is an expertise in project management and quality control you need to grow to succeed in the long term. And if you still have doubts: all the startups that went through HAXLR8R learned how to do it, and so can you. It might be worth noting that our 30 startups work with 30 different factories, to suit their needs best. To each his own!

“Not every product can be made in a toy factory. Your factory will make or break you, so pick wisely. Do not have a third party between you and your key suppliers.” — Zach “Hoeken” Smith, Co-founder, MakerBot (HAXLR8R Technical Director)

Be Memorable
Good products have good technology and design, but also good distribution and good branding. Can your brand encompass your values, mission, product and be memorable in many languages?

As an example, Axio, a company doing wearable tech for your brain and part of the second batch at HAXLR8R, did a thorough rebranding to Melon, revisiting entirely its name, logo, approach, app and look and feel.

Pick the right business model(s)

Pure hardware is at risk of being commoditized fast, and many products now include “smart” connectivity and a service layer. Doing hardware-as-a-service means that picking the right business model is very important.

At HAXLR8R, Spark started off by making open source hardware and software, then a device-based subscription and then licensed both hardware and software. More recently, Vibease, the wearable smart vibrator, added sales of the content of a library of audio recordings to its revenue models.

Changes in the ecosystem have made it much more possible to start hardware companies at a much lower starting cost. Expect to see many new successful hardware startups, as “smart” products gradually come to replace the objects in our life!