Previous Submission

Failing at Hardware

Next Submission

Most of the startups we think about can fit quite neatly into the asset-light, fast-iterating, lean mold– easily pivotable from spin to spin of an idea until stakeholders’ patience runs out (ala Yabbly). Hardware startups have a lot more to worry about — such as the much greater difficulty of creating functional prototypes, iterating, translating from prototype to production, and then the whole range of traditional logistics challenges (lead times, demand forecasting, manufacturer relations, packing, shipping, warehousing, customs duties, etc.), the last of which is often the most problematic. Failing fast, failing often thus seems a lot more challenging for hardware startups, and the high-profile failures of several crowdfunded hardware projects has shone a spotlight on the challenges involved.

Yet, interestingly, many of the lessons that founders of failed hardware startups list are similar to those we see elsewhere. The founder of KOLOS, a crowdfunded iPad accessory for racing games, for instance, listed the following as his top lessons:

1) Scratch your own itch

2) Make sure there’s competition

3) Build an MVP

4) Talk to customers

5) Don’t ignore the product

6) Avoid tackling too narrow of a niche

7) Take a good look at the market and its potential

8) Spend your funding carefully

9) Don’t be naïve

10) Fail fast

These are all general lessons, applicable to most startups, yet the relative weights are perhaps different for hardware. #1 and 5 are critically important, as hardware tends to introduce complexity that makes it more difficult to change the basic idea, and thus a resolve and investment in the idea (#1). Hardware also necessitates extra creativity to come up with a useful MVP (#3) for testing purposes — this is tricky because the surface details that differ between an MVP and a later version may be the very things that matter most to users (how would one usefully MVP a new smartphone, for instance, beyond a styling buck?). #8 becomes important because of the inherent cost and time bloat that accompanies the space.

Hardware doesn’t present an entirely different ballgame– the same principles we normally discuss for software or other areas still apply. Yet the differences necessitate additional planning to accommodate the lesser flexibility for later changes.

2 thoughts on “Failing at Hardware

  1. Excellent points here! The recent proliferation of hardware start-ups is both exciting and worrisome to me. I read an interesting case on a new hardware “accelerator” in Boston called Bolt (https://www.bolt.io). Bolt helps companies progress from an idea to a prototype to a product on hundreds of shelves. Historically, any company hoping to manufacture hardware had to be well known, established and have a lot of money. A common refrain within the tech industry was and still is “hardware is hard” because it quite frankly is. Many high profile companies have failed, in part because they weren’t able to navigate the complex world of manufacturing and supply chain management. Thanks to companies like Bolt, this is no longer the case (or as much of the case) today. While this may seem wonderful at first glance, I can’t help but wonder if it’s not and if we are just tampering with the principles of natural selection. As the cost of starting a hardware businesses has fallen, the number of hardware businesses has risen. As the opportunities in this industry have risen the risk profile of the businesses within this industry have risen as well. This, in turn, implies that the number of hardware startup failures will rise as well. This rise in hardware failures is precisely why I feel a sense of trepidation about the changes taking place in this industry and also why the points you’ve mentioned are incredibly relevant ones. Thank you for sharing!

  2. Great post! I think two additional unique challenges that hardware startups face are related to hiring and the lengthy product development lifecycle. Hiring technical talent is tougher than in software startups because (1) often, you need great mechanical and electrical engineering engineers, in addition to software engineers, and (2) there’s not a tonne of hardware engineering talent out there. Also, the lengthy product development cycle makes it harder to pivot from mistakes. One approach that hardware founders often adopt is to include as many commodity components in their design as possible to reduce cost and avoid the capex involved in setting up an in-house machine shop.

Leave a comment