How difficult is it to build hardware for a startup?

My gosh… I’ve always been a software guy but since I’m young I like to build things. It starts with lego, even “Mecano” (ok I’m not that young anymore and I have played with Mecanos as a kid..) and last year I though it would be really fun to try to build some kind of hardware and see how that works. I was inspired by the Makers movement and all the Arduino’s projects and thought to myself: I can do this!  So I did..

Here was the plan:

  1. Find a good idea (I had one already)
  2. Build a prototype
  3. Put it on Kickstarter and raise some funds
  4. Sell it and potentially expand the product line

That sounded about right.. My initial goal was to learn about hardware making and prototyping.. All those Kickstarter projects were pretty fascinating and the advent of 3D printing allows you to go pretty far. Hoping to help (or warn) other entrepreneurs about going that route, here is what happened:

The context:

I watch TV and I have kids.. I can never watch a good action or sci-fi movie without the risk of waking up the kids or annoying my neighbors so I wanted a device to wirelessly transmit sound from my TV to my ears. There are plenty of them out there already but without going into the details, they SUCK! The only ones that work very well are the Sennheiser, which will charge you an arm for them and force you into a single type of headset.. I have my headsets, some nice Bose QC15 that work perfectly and I wanted to use those instead.

The Idea:

Have a very small device on which I could plug my own headset to listen to the sound of my TV wirelessly.

The constraints:

Wirelessly transmitting sound is not easy when combined with TV: It cannot suffer any delays (latency) between the image and the sound otherwise it would look like a horrible foreign language dubbing! And it cannot use too much bandwidth because there are a lot of interferences with existing wifi networks already. Wifi is out, Bluetooth is out, regular RF is out and the only really good solution I found was the KLEER chip (also used by the high-end Sennheiser system)

The easy part:

Get a name: Soundora.com was available..

Get a quick website up and running www.soundora.com

Get a Logo (I have some really talented friends there)

soundora_high_transparent

Get a twitter account, an email and a facebook page

Barely any costs there.

Building the first prototype:

I found who built the KLEER chip, and luckily they were in the Silicon Valley so it was easy to contact them. They send me all the specs of the chips some reference designs, and all the surrounding hardware.. That’s where the nightmare started.. My electrical engineering skills are about 15 years old and those chips cannot be soldered by hand.. Half of the surrounding component I had no clue what they were, some had different options I didn’t understand either..

So after a bit of research I reached out to companies that built prototypes for you. They know their stuff! Great conversation but the initial investment was at minimum around $10K to have a single prototype, and probably around $50K to get a first low volume batch.. No go.

The fun part is that I initially did some designs without hardware constraints, and had a good idea of what it would look like, but as soon as you add: Battery, Connectors, and all the other stuff around the device, you realize the amount of magic required to build something as small as an iPod Shuffle is! Man, Apple is good.. So all my initial designs went to the toilet. I had to start from the inside and work my way out.

So I bought a couple of low end Sennheiser headset (which use the same KLEER chip as the high end) and tore them apart . I took the board and put it inside small box I bought at Frys.. What I didn’t expect is that buying components such as a 3.5mm jack plug, flat batteries, etc.. would take so much time. I also didn’t realize that they were never exactly the right format, the right length, etc.. Amazon doesn’t have a lot of it so delivery would always take 5 to 7 days to realize it wasn’t what I needed.. In the end I did get all my part and a bit of soldering later I had a working device:

First prototype of Soundora

When I say working, it’s partly true.. The button in the front for power and Volume were iphone buttons that are way to thin to work with.. I could not really change the volume or turn on and off the device.. I had to design my own box (which was fun).

Here you go a few days later I had my 3D model:

3D model or Soundora receiver

I used Shapeways.com which had the cheapest 3D printing service I found and got me the box. Very existing but another few weeks to get it..

1 IMG_8225  3D Printed receiver open

 

At this stage I realized that 3D printed is not really there yet.. Anything under 0.8mm space will be filled.. That means that my button did NOT work (again!!). I was starting to get frustrated with this, buttons are really tough to design, it’s just amazing!

Anyway, I got the circuit in and kicked it up

2 IMG_8228 2 IMG_8233

 

huh… That orange light wasn’t supposed to show up that much.. I even had build a little deflection inside the box to let the light come out through a small hole.. Interesting surprise.

Anyway, it worked! I had to have another batch of buttons from Shapeways to solve my issued and it now looks pretty cool.

I also found another battery pack that is flatter and should work as well so I was in good shape. I’ll have to do another box (thinner) though.

But here is the main issue: Ripping out a Sennheiser device for my personal use is fine but I can’t really do that commercially so I still needed to find a way to get my own circuit built.

Getting my own:

By total chance, I stumbled upon a small business that is building systems around KLEER for different customers.. They built the Dr Dre wireless headset.. so I contacted them. Very cool guys and they had small circuits (smaller than sennheiser) that I could use to get an even better device. Very exciting.

But then we talked about production batches.. The Kleer chips are bought by batches of 2000, pretty expensive per unit cost.. Building the box (transmitter and receiver) would probably cost about $4K to $5K to get the mold and get it ready for a batch. That doesn’t even count for prototyping the real thing..

Basically I need to raise at least $100 to $150K to get the project going, and it would mean a sales price in the $150 to $200 a system (1 transmitter + 1 receiver). You can add more receivers (up to 4 total) but it cost about as much as the transmitter.. So the economics barely work for small batches of a few hundred devices. They would for larger batches though.

I am currently kind of stuck there, I don’t want to do a Kickstarter project with a prototype that doesn’t look like a finished product (maybe I’m wrong in thinking that by the way), and I am honestly having a hard time with anything touching the electronic of the device.. I’m a software guy. Also I want to keep it lean as much as possible but I realize that the lean startup model of software doesn’t apply that well in the hardware world.

Other things to consider:

Even if I get to a point where I do have a working prototype there are other things to think about when building a hardware startup:

  • This one is a wireless device and requires certifications about the lack of danger of the transmitted wave. There are a bunch of certs you need to get for the device that will add costs to the project.
  • Designs should be turned from the current Patent pending status to a real patent in some way even though it’s using only existing technologies.
  • Packaging is another good one: Getting a nice box that is cheap but looks good and conveys the experience you want to provide to the users
  • Power plug adapters for distribution in Europe and North america, including different voltage and frequency
  • Documentation in several language
  • Distribution network: initially direct as done through kickstarter, but eventually through stores like Best Buys and others. This will hit the margins.
  • Shipping costs between China (or wherever the device is manufactured) and the destination
  • Maintenance and defects management: Some devices won’t work, some users will be unhappy, and you have to determine who is responsible, what’s your return policy, what costs can be projected around those issues, including shipping..
  • Potential competition from Sennheiser or others.. Roku came out with a similar device and even though it is limited to one listener and to sources from the ROKU box, it still is a fairly tough competition.
  • Unsold devices: If you create batches of devices, what if you don’t sell them all?
  • Delays in production and shipping..
  • etc..

Conclusion

I didn’t give up. I still would like to get to a point where I have a very nice looking prototype, but when I compare this to any of the software project I have done, I realize that it is a totally different world.. Fun in a way as you can touch and feel something, and as you are effectively building something, but the initial cost and technical skills required are really larger.

I have learned so much along the way and I now look at all those Kickstarter project in owe of the complexity I lived through. Some of those guys are really good.

I hope you you entrepreneurs out there will find value in this experience and maybe make you think a bit more before going into hardware. I also welcome any help from someone with electrical engineering, manufacturing, hardware engineering to get this through a first finish line 😉

Cheers, and keep making stuff!

Advertisements

13 thoughts on “How difficult is it to build hardware for a startup?

  1. Great Post and good luck!
    BTW – If you choose an existing technology for the transmission, doesn’t it already come with the certs?

  2. Buddy,

    Great post, thanks for the effort!
    I feel your pain and can only imagine the hassle of prototyping hardware.
    Reading through your story, it looks like there’s one aspect you didn’t mention and which is of upmost importance (particularly in the lean startup model): have you talked to the customer?
    Finding 5K, 10K, even $200K is not so tough once you have demonstrated that your main underlying assumption -people actually do care about your product- is true.

    Your idea seems good at addressing your own problem, but how many people face the same frustrations?

    Isn’t there a cheaper way than a real (but necessarily not as nice as a sleek Apple product) prototype to test that very assumption?

    Building a working prototype induces a strong bias in your testing experience: people will judge its design. They will compare it with that of your competitors, or to their usual references, say iPod, etc.
    They’ll give you answers to questions you didn’t ask:
    – how cool is this product?
    – can I brag about it with my friends?
    – etc.

    Building a sleek and sexy prototype will cost you an arm and a leg. And it still won’t help you answer your key question: does my product solve others’ problem?

    You probably have thought about all this already, but I know that sometimes you’re so excited about your idea that it’s easy to get blindfold and forget about the basics.

    Who are your target customers? Can’t you find enough of them to have a live discussion and judge their perception?

    If people are indeed interested, how many of them are out there? How much would they be willing to pay? How will you reach them?

    Is your idea patentable? Can’t you just patent it and sell it to a major player who will have the scale to make it affordable to the masses?

    Available to discuss this whenever you want over Skype 🙂

    Congrats again for bringing it so far already!

    • Hey Francois, I definitely agree with you on talking to customers first. I totally did talk about the idea to a lot of people and I got extensive positive feedback about the need and usage, even had some surprises like one guy telling me his wife was not happy when he was playing video games in the evenings and waking him up, which is an entire market segment all together that I hadn’t thought about. The willingness to pay is something I haven’t assessed very much yet though as my idea was to get a prototype and get it in the hands of people to try it out and get me more feedback and address the willingness to pay from it. I think people may not entirely realize the benefit of the solution until you try it out. Many of the folks I talked to who had used similar systems were saying that they got addicted to it but the quality and usability was so bad in existing products that they were frustrated.
      I did get one strong push back around having an extra device on you and wanted to use Smartphones instead which I found really interested and drove me to look a lot into this option without success. The available technologies in smartphones cannot work for TV unfortunately.

      Also, keep in mind that this project was also a way for me to learn about building a hardware, which could be totally de-correlated from having any market to address.

  3. great insight and “behind the scenes” kickstarter stuff. So I guess a LOT of projects on KS are just not even at real prototype level and people don’t realize how tough it is to build actual electronic devices.
    Don’t give up !

  4. Hi, congrats for pushing forward this project, I really enjoyed reading and helped me look at hw startup projects in a different way. Good luck with the rest of your ventures!

  5. After checking out a handful of the blog posts on
    your web site, I seriously like your technique of blogging.
    I book-marked it to my bookmark webpage list and will be
    checking back soon. Please check out my website as well and
    let me know your opinion.

  6. Hello Antony,
    i need kleer chip for an application. But i did not find Kleer Chip. Can you give me a link or website adress for buying Kleer chip. Thank you!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s