Wednesday, December 9, 2009

Lesson #1 - Hardware is Hard

Among my first lessons learned at Zeemote, Inc. is that hardware is hard and in my opinion start-up companies should avoid getting into the physical product business if at all possible.

For some time I believed that physical products held a huge advantage over software-only products. Consumers are after all used to paying for items that they can touch and feel whereas most software these days seems to be free. The high barrier to entry of the hardware market also ensures that you will see much less competition than you otherwise might.

Having said that, I now feel that the disadvantages of creating hardware products far outweigh the advantages. The barrier to entry is high for a reason and it is not just your competition that needs to cross it!

Compared to software, hardware is several orders of magnitude more expensive to prototype, manufacturer, and distribute. One software engineer can develop a basic demo in a matter of hours and the manufacturing / distribution costs approach zero. Creating and distributing a hardware product requires much more expertise across a wider range of domains. It requires hiring more staff and partnering with outside (and expensive) vendors. An initial prototype may not be seen for months. If you make it to manufacturing, each unit you build costs money and everyone along the value chain needs to get paid. You must figure out how to distribute the product and manage all the logistics it entails. You also need to carefully manage inventory which is non-trivial since forecasting demand of a new product is almost impossible.

The larger (but related) issue is that of iteration. Start-ups are usually attempting to create a new market or segment an existing one. A start-up is usually founded on some hypothesis which almost always turns out to be wrong. Experienced and successful entrepreneurs will tell you that the key to success is through experimentation, learning and iterating on your product idea. Get your product into the market as quickly as possible and modify it based on what you learn. Often the final winning product bares little resemblance to the initial idea. The sad reality is that today the costs of manufacturing and distributing hardware makes iteration impractical for most. If you don't get it right the first time then you are dead meat.

Unfortunately at Zeemote we spent much more time dealing with the basic issues of just bringing the device to market than we did (or could) on iterating the product based on user feedback.

If you are considering a hardware product I'd urge you to think long and hard about it. Is there any way that you can develop an initial version or a similar product that doesn't require hardware?

I think a good example of this done right is Boxee. They provide an internet media streaming service that ideally would run on a dedicated set-top box connected to your TV (like your cable box). When Boxee launched however they did so with desktop software only (no hardware). This let them test their product idea and build a user base. Sure, ideally, it would be a set top box but the software allowed early adopters to try the service and provide feedback. It wasn't until this week that they announced a dedicated hardware device. Smart move.

I don't see myself getting into the hardware business again any time soon but if I do you can be sure it will have a software-only strategy at launch.

- mpv


Elaine Chen said...

Yep, hardware is hard. Lean startup/MVP concepts apply a whole lot better to software. Hardware is a much more capital intensive play and the risks are way higher while the rewards may not be that much higher. That said, it's not helpful to advice founders to avoid physical products if the idea they are passionate about requires a piece of hardware. If they see a way to do it and can accept the risk, I say go for it anyways. What's a startup without a little risk?

mpv said...

I agree, which is why when I used the weasel words "if at all possible" when saying to avoid hardware. Lots of products require hardware (like the Zeemote) I think its just harder to get a hit since its so much more expensive and time consuming to evolve the concept.

z said...

Sounds pretty familiar...

mpv said...

Wow. Lots of similarities with fitbit!

Elaine Chen said...

Fitbit's problem is taking preorders and then failing to execute / fulfill. Zeemote's problem is either selling a flawed product or failing to sell in / sell through... not quite the same problem in my book.

z said...

Entrepreneurship will always be characterized by taking risk. It's the nature of the beast. If you're the entrepreneur, though, you want to make sure the risk is as calculated and manageable as possible. That's where I think Mike's advice could potentially be useful to someone. I seriously doubt any passionate entrepreneur will be dissuaded from following his/her dream of creating the next thingamadoodad because Mike advised against it; if that does happen, then the product probably isn't ready for market anyway. But perhaps Mike's experience can help a new enterepreneur better plan his/her strategy and put the whole effort into a bit more perspective.

Anonymous said...

I want to quote your post in my blog. It can?
And you et an account on Twitter?

mpv said...

Sure, no problem. Post a link here though. And my twitter name is "mpv". :-)