More on defensibility

Alright, I admit that I did not really talk much about defensibility in the last piece, despite a title that suggested so. I hereby (try to) compensate for that so bare with me here.

Defensibility for a startup refers to the ability to stay ahead of the crowd after successfully ramping up a new service or even creating a new market. While the definition of the term hasn't really changed, its means has.

I remember when me and my former colleagues started our own semiconductor startup in 2004, worries about patent infringement had always been at the back of our minds. Technological advantages were and still are so important that startups are at a natural disadvantage to big firms, who've over the years amassed huge amount of patents.

And as I pointed out in the last piece, patents only function in pools. Having a single great patent is useless because goliaths like IBM or Google will be able to retaliate with other patents even if they lose out on your single, brilliant patent.

Patent war have really evolved into what the name suggests, a war. In a war you don't send out your best cavalier asking for a duel — at least not since the end of the romantic middle age — you square off with your enemnies by organizing your troops of different functions into a system. Before full confrontation, both parties would already have established an idea of the odds of winning the war. Some skirmishes might be triggered deliberately to test out own theories on the real prowess of the other side, but ultimately the best solution is to strike a peace treaty which includes mutual compensation that could avoid full confrontation.

To anti-quote Robert Downey Jr. in the stupidly entertaining The Avengers: you have a Hulk (a single patent), but we have an army (a patent pool)!

This was why in the go-go 90's, semiconductor startups that made it, like Broadcom, acquired new startups as fast as they could so that they could get the engineering talents, sometimes the products but most of the time the patents.

Even non-semiconductor tech firms do so regularly, once they complete their IPOs and are awash of cash — or better, are operatingly cash positive. Google pulled the famous trick of a Motorola flash sale by retaining the patent pool while dumping the hopeless product lines and company assets. And do not be fooled by the media discussing one or two particular patents — a multi-gesture touch panel, blah-blah-recognition-system, etc — in the lawsuits between Samsung and Apple, what's really behind those patents that were singled out were two huge patent pools. Without the pools, neither would have even entered the stinky patent battle.


The patent pools were so much of a concern in Silicon Valley that before Facebook went IPO in 2012, IEEE Spectrum Magazine — the equivalent of Vogue or GQ magazines for electrical engineers in the States — ran a piece comparing the then-already-giant Google with the up-and-coming Facebook, in which one glaring gap between the two was the patent numbers. I do not have the archive at hand but suffice it to say that Facebook had ridiculously small numbers of patent on the eve of its IPO, both granted and pending. The article went on to suggest that this was yet another sign that Facebook wasn't really innovative and wouldn't be able to defend itself facing competitions from both Goolge and new startups.

Except that that has NOT been the case.

Fast ward by more than two years. No one would question the defensibility of Facebook as a social network. Not only were it able to continue to grow its revenue, it also addressed properly the concern of not being mobile enough on the even of its IPO. The trajectories of the Facebook revenues and profits over the years are so classical that Wall Street saw it as a sign of predictable future and pushed Facebook's market cap to over $200bn.

How has Facebook been defending itself? Certainly not with patent as I can't really recall any significant patent lawsuits involving the social networking giant. And even today I would still argue that it's not an innovative company, at least not on the user-facing front as we pretty much still do the same simple things on Facebook as we did 2+ years ago.

Its defensibility lies in the simple networking effect — if you and your friends are all in the system, then it's hard to switch to a different platform, however much you hate the existing solution.

So you say: blah-blah-blah, everybody knows that. So what?

Well, if patents are not the best ally of a tech startup in war, then that leaves execution as the only viable one, as I discussed in the last article. For any startup offering that quickly becomes popular and grows by itself, there are bound to be lots of copycats, simply because that the ideas that catch on fast with consumers are also usually the simpler ones, which usually imply less complicated technologies involved, which in turns invite copycats to join the party.

However, if you focus on execution — enhancing the user experiences specific to your offering, growing exponentially while maintaining system reliability, addressing user concerns by updating functions, expanding geographically to strengthen the networking effect, carefully building an own eco-system and eventually achieving a recognizable brand. Here the defensibility manifests itself in multiple factors, making it even harder for your copy cats to catch you.  Your copy cats would have much less room for maneuvering and would be most of the time behind the curve, while you would be the one that grow organically and the distance between you and the copy cats would be larger and larger. The next thing you know, your hand is on the button of the IPO ceremony at the Nasdaq floor.

"What if the copy cats simply engage in price wars with me?", you ask. Shouldn't patents be useful then to forbid them to do so, say, asking the court to issue an injunction?

As I said in the last article, as a startup you don't have the time and energy to engage in patent lawsuits, not to mention those against equally cash-deprived startups (and hence less or zero compensation even if you win). If simply by triggering a price war a copy cat could grab your market share more than temporarily, that only means the market is not worth your time and effort. Recall in microeconomics class that an ultimate price war exists in a complete competition environment where every supplier makes zero profit. Some such sectors might justify its existence by exhibiting very low risk, e.g. bakery shops or tabacco shops, but these are no places for highly risky startup endeavours that walks around with capital priced at 30% IRR.

In short, you shall then thank your copy cats for proving that the market isn't worth your effort for you — by using their own resources. And you shall get out of the sector quickly, or use the Silicon Valley buzzword, pivot!

5 lessons I learned as an entrepreneur

Amazon acquiring Twitch and thoughts on pivoting