Monday, October 26, 2009

Why do you do Hardware?

I was asked this question by a young VC analyst replete with Stanford Biz school education, attitude, and three months of on-the-job experience. It struck me that perhaps the industry has passed me by and Hardware is so 80s ... so I thought it was time to soul search as follows: the rivalry between hardware and software is akin to that between physicists and chemists. Many would say that when you know what you are doing you do it in hardware and if you don’t [know what you're doing] you chose software ;-)

At Iridex I was stunned at the power of software, particularly with the advent of the touch screen. Suddenly we could change the bezel with code only, we didn’t need to retool or repaint the knobs. It was also incredibly more resilient in nasty places like the OR (well it wasn’t at first but it eventually got there). The difference from the hardware guys is that you can afford to get it wrong because it's easy to fix (compared to messing up the mould for the bezel or display, which could take months to fix). However, to hardware guys, the software is always the gating item in a project because it's always late, and never works properly--in fairness the software can’t be done until the hardware is complete so it’s a shared responsibility for being late.

The logic of doing software deals is that the recipe is fairly well defined. Because there is no manufacturing you can usually see the product or a prototype before you invest, talk with beta customers and a have a clear understanding of the value proposition--but where is the fun in that ;-) In hardware, it's rare to have the product, particularly if it’s a chip, so you have to rely on market vision, trends, and team. The capital expense is much higher in hardware and if you get it wrong it's really hard to recover. On the positive side, hardware begets a lot of IP and if managed properly you will always have a residual value to your investment even if the market fails. This IP barrier is the critical missing piece for me in software deals, unless there are legitimate ways to protect algorithms (like crypto, sequencing, expert systems, etc.). They tend to become execution plays, with tender hearted companies like Microsoft and Google looking over your shoulder ready to step in with a mediocre purchase price based on a make/buy decision. Of course if you can execute, very large companies can be built, which is why there are so many software-centric VCs.

In contrast chip investors seem to be a dying breed--partly because VCs don’t have the firepower to fund $100M deals ... well, unless it's in Cleantech ;-) Yet in medical device or biotech these $ amounts are fairly typical, so what's going on there?

In the medical space, in addition to wonderful IP barriers, you also have the FDA as a blessing and a curse. On the bad side it can take five years to get approval from scratch, but the value of that asset is immense, almost regardless of revenue metrics. To a large company like Medtronic or J&J, it's well worth paying $100M for a “failed” device company that has the platinum of FDA approval--often failed deals turn into great successes when big pharma puts its marketing muscle behind them.

Virtualization and cloud computing are strong trends and important ones--but as a hardware guy I worry about the security of these connections, and the power of software to effectively secure these open network systems.

I have seen a new breed of software guys who make software hook into the silicon itself, which turns out to be a wonderful way to get chip guys who don’t do software to appreciate it, and leverage the hardware to work in concert with the software to make a secure device. I think this is probably a theme.

I guess in the final analysis no matter how virtualized you are, at some point you have to touch the real world so software is pretty useless without the hardware that manifests it, so to me at least they are inextricably linked in a timeless symbiosis ;-)

Responses to comments:
David Wright - Larry - I have had a few successes and one deemed a failure. To me it is beyond doubt that you "can" learn a lot more from a failure than successes - how far can you push yourself, your team what are the warning signs etc but I don't think many will count the scenario you presented as a failure. Timing also is an issue. In my case - I am a better CEO because of my company failure and what I learnt to get it to the heights achieved but this was permanent - the company folded so a genuine failure. Until this happens the scoreboard is still running - will this person still consider it a failure if that company turns up again. If not they should tell Mr Murdoch how much of a failure he is :)

David--Thanks for your thoughts. We invariably learn more from our failures than successes--and how we react when things don’t go as planned really teaches ourselves and others about the quality of our character. Sounds to me like you will have continued success despite some setbacks.