We live in an age of stunning technologies that defy the understanding of most non-specialists. And we live in an age of sound bites and short attention spans. These two things can come into conflict when it comes time to raise capital. The overwhelming urge -- even the necessity -- is to boil high concepts down to easily digestible sound bites and drown out all the nuances. Thus, salesforce.com was once "the end of software" before it somewhat ironically became "software-as-a-service"; Renren is "the Facebook of China"; and we are always on a search for "the next Apple."
Pacific Biosciences of California (Nasdaq: PACB) has gotten a lot of lift from its own sound bites, too: It gained a great deal of notoriety a few years back by claiming it would be able to produce a complete, high-quality human genome in 15 minutes by 2013, for a cost of only around $1,000. These claims -- backed by some truly virtuoso science -- helped the company land $370 million in venture capital and another $200 million from its October 2010 IPO.
But in April, PacBio made the transition from a company selling a great scientific story to one selling real machines. And the reality is that the 15-minute genome, much less the $1,000 price tag, is still a long way off. Concepts the gene jocks once drooled over are now, in some corners, getting rounded off to "hype." As a recent Businessweek article put it, "The consensus view among rivals and some outsiders is that the company is unrealistic about its commercial prospects."
CEO Hugh Martin put it to me another way when I visited the company this week. "How do you raise $370 million as a private company? You raise $370 million saying, 'I want to show you where we can go.' So that's what we did." Yet Martin says the company isn't backing off its original claims about the stunning gains the company will make in the speed and cost of sequencing (though he is careful to avoid any comments about timing).
"We know how to do this," he says. "But we had to get to market quickly. And so once you have something to sell, the last thing in the world you want to be doing is waving your arms about [what the company can eventually achieve]; you want to be waving your arms about what you're selling."
That change of focus and some early stats on the machine -- notably its relatively low 85% single-pass accuracy rate versus a similarly priced HiSeq machine from Illumina (Nasdaq: ILMN) -- has given doubters a lot of ammunition. That's a shame, perhaps, for investors who bought into the IPO, as they've seen the stock price crumble about 30% from its debut price of $16.50 per share. But skepticism may prove a hidden boon to patient investors who are looking at the stock now.
Disruptive technologies don't usually appear fully realized. Indeed, in The Innovator's Dilemma, Harvard business professor Clayton Christensen declares that real disruptions often begin with tangibly inferior products that can't compete in existing markets ... and so create new ones. Think of, say, solid-state flash drives that once cost so much more than standard hard drives they couldn't possibly hope to compete in the PC market. They instead found a home in things like cameras and then other mobile devices before finally showing up in laptops.
That's not to say that PacBio is offering up an inferior technology -- far from it. If anyone ever told you to not let the perfect be the enemy of the good, this is a case of not letting the world-changingly jaw-dropping be the enemy of the absolutely incredible.
Martin argues -- pretty convincingly in my opinion -- that the 85% accuracy number is a red herring. That's because, he says, sources of error in the system are entirely nonsystemic. In other words, sequencing errors show up in random places and are highly unlikely to appear in the same place on a second pass, meaning multiple passes (that is, more coverage) will quickly arrive at an extremely accurate result. And the machine is so fast that multiple passes (which are standard practice in all gene sequencing) can be quickly accomplished. Because current pyrosequencing methods rely on things like DNA amplification, errors may be fewer but more persistent, so even multiple passes could give you the same erroneous result.
That's a tough argument to make to investors -- there's that PR problem again -- and perhaps even a tough argument to make with some users. But data will eventually convince PacBio's customers, even if Wall Street lags behind. Importantly, Martin notes, the company can improve its speed and cost an order of magnitude or more with the current system, through advances it is making with reagents and the disposable chips used for the system. That's an important consideration for any lab looking to plunk down $695,000 on a device it doesn't want to see become quickly obsolete.
Nevertheless, users are interested in what PacBio can do better than its competitors today. The company isn't initially going to go head-to-head against industry leaders Illumina and Life Technologies (Nasdaq: LIFE) when it comes to whole human genome sequencing; it's going to start in niches where it can accomplish something harder to achieve with other technologies. That will mean sequencing tricky regions with lots of repeating nucleotides, or even working in parallel with other methods to help assemble fragmented genomes.
It will mean sequencing pathogens, which gives the company a particular edge when time is of the essence (something the company proved by sequencing five strains of cholera from a outbreak in Haiti in mere hours, the results of which were subsequently published in The New England Journal of Medicine). Perhaps most significantly, the PacBio RS offers great potential in cancer research and diagnostics. The company can leverage these present-days advantages while the technology continues to improve.
These days, Pacific Biosciences is hard to break down to a sound bite, which may explain why some investors seem to have decided the technology isn't and never will be ready for prime time. But long-term investors willing to accept some risk might want to remember that this is exactly how disruption takes place.
