Creative destruction in the personal genomics industry?
I’m hearing about rumblings at 23andMe, and not in a good way. The company made a big splash a few years ago, and came highly recommended by friends (e.g., “They know their science, and have a bottomless pool of money”). This story at BNET got my attention though, and confirmed what many have been hinting at, or just telling me straight-up. Let’s start the from the beginning. Back in late 2008 23andMe seemed absolutely untouchable. Here’s Andrew Yates of Think Gene from then:
People, 23andMe isn’t going anywhere. They are the Bill & Melinda Gates Sergey & Anne Brin Foundation, Silicon Valley style. Anne Wojcicki is married to Sergey Brin, so 23andMe has access to all the talent, connections, and capital 23andMe would ever need to make 23andMe work. Thus, assuming 23andMe doesn’t do anything egregious, they will exist for as long as Mrs. Anne Wojcicki Brin pleases it to be so. If 23andMe shuts down, it won’t be for some mundane reason like the bills weren’t paid, it will be because Anne felt like it.…What, does that offend your meritocratic, democratic, American dream sensibilities? Too bad. Go get an Ivy+ degree and marry your own richest man in the world.
This seemed to be a reasonable assessment, and I know I shared it. Case closed.
And the momentum kept going for a while. But since then something seems to have gone wrong in the narrative. Check out this this comment thread at Dr. Daniel MacArthur’s. There’s apparently some lack of clarity about why Linda Avery, the co-founder, left. Who would want to leave a company which had access to Sergey Brin’s pocketbook?
As a friend put it to me, what kind of co-founder leaves a successful startup mid-stream voluntarily? (excepting Jawed Karim) You can get some ideas why from the Glassdoor reviews:
Employee #1 (Feb. 3, 2010):
All decisions came down to one person, and it was NOT the ceo. Senior managers made some grave mistakes and many employees suffered. Managers and co-workers inexperienced. Did not value customer satisfaction. Vague product definition with very little marketing support.Employee #2 (Nov. 17, 2009):
Extremely poor management that completely ignores the customer as well as employee needs. Major legal threats to the very existence of the business. Poor pay and complete lack of communication. Lack of focus and direction in management decisions.
This one is extremely well informed and presages the second round of layoffs:
Employee #3 (Oct. 1, 2009):
23andMe’s non-traditional approach is a strength in building new and interesting products and in attracting a very savvy group of customers. 23andMe has done a lot for genetics–making a complex topic accessible and fun, bringing genetics to a wide audience, and also influencing important discussions at very high levels.The non-traditional approach is also a weakness, because the company’s lack of corporate biotech experience at nearly all levels of the company has created major blind spots. As a result, 23andMe has stumbled into barriers that a more experienced company would have been aware of and taken the necessary steps to mitigate–which can require major commitments in resources and incorporation into the long-term business plan.
23andMe also made an early choice to pursue several different, non-overlapping goals at once, and although it has achieved some success–excellent content, unique website, novel research platform, recognition in the space–continued growth will start to cause these functions to overlap less and less, potentially creating resource conflict and mission creep in the future.
The lack of biotech and research experience in the executive management and BD teams also creates an attitude of taking research for granted. Biology is hard; finding discoveries that can provide a foundation for profitability is by no means guaranteed. Improving the chance of medically useful discoveries means thoroughly understanding a clinical question and conducting the best research possible–which may directly conflict with making the web service as customer-friendly as possible. But the lack of direct research knowledge and experience has led 23andMe to avoid focusing on specific areas, which means that specific expertise has never been brought on board and that research quality continually compromises with (and is compromised by) the needs of the web-based service.
My advice is to build a company that focuses on doing one thing very well. If that thing is a web portal for genetics content, then set aside the research goal. If that thing is research, then set aside the web portal. If it is a genome-wide diagnostics company, then do whatever it takes to succeed in the increasingly difficult diagnostics world. In terms of both science and profitability, quality scientific research requires a much larger commitment of time, resources, and discipline than management has been able to show thus far. Hiring people with clinical knowledge and with experience at running a successful biotech or molecular diagnostics company will keep 23andMe from having to constantly reinvent the wheel.
Note these reviews were anonymous. But now we hear about the big pay out. When we heard about the lay offs last year (thanks to Dr. Daniel in my case), we didn’t know that 23andMe also spent $4 million dollars in payments to an executive, as this SEC filing shows.
The likelihood is that either Linda Avey got $4 million in severance to keep quiet, or Anne Wojcicki (wife of a billionaire) was taking $4 million from 23andMe’s coffers while almost 30 people are fired. 23andMe’s PR states that this was paying back a “loan” from Anne. OK. Let’s keep it real, for the wife of Sergey Brin (net worth around $18 billion), $4 million is pocket change.
The only thing I can think of is that Sergey Brin got tired of throwing good money after bad. A few years ago I blithely suggested to Dr. Daniel MacArthur that deCODE would always have access to Icelandic money because of national pride. Well, that didn’t work out, did it? I also mentioned to Dr. Dan that Brin would always be willing to pump money into 23andMe, but now who knows?
Word is that morale at the company is at an all-time low, with Patrick Chung’s layoffs decimating 23andMe’s invincible image. And the hits may just keep on coming. Linda Avey noted on her blog that Andrew Pollack from the New York Times is about to write one of his characteristic brutal hit pieces. If so, it’ll be a big turnaround for a company which was initially showered by media praise, and assumed to be untouchable.
A hard question you can ask a start-up is, why would anyone want to buy your product?
In the case of 23andme, the medical diagnostic applications aren’t very convincing. For many of the diagnostic tests they do, there are too many false positives. (Base rate fallacy, etc)
The “family history” angle looks more promising—discover from your mtDNA you have Ashkenazi Jewish ancestry you didn’t know about; get in touch with your fifth cousin who was adopted as a baby, etc.
There are some serious legal and ethical aspects with both aspects of the service, of course.
There’s an almost dream-like quality to the story of how this particular company came into being, as if it were designed as literary device in a novel about life and love in the hay days of silicon valley but underneath all that a lesson of how guaranteed success drives us to unhappiness as if the gods see to it, for their entertainment but to serve as a lesson to the rest of us monkeys. Sounds more like soap opera than corporate history, though it’s all about human drives, feelings, aspirations for power and control, and there will always be the voyeuristic aspects to watching someone cast their personal fortunes onto the most promising water, and waiting for fish to rise…big fish…before the bait falls apart or is nibble away by small fry.
I wouldn’t bet that 4 million dollars feels like small change to people just because they have far more than they can spend. Our human sense of wealth seems to be related to having some notion of what we can do with it and not just how much we have left, because there can never be too much, let alone enough.
I am a totally dissatisfied customer of 23andMe. In fact I have asked for my money back. I paid a lot in my Australian currency for a service inadequately fulfilled. I have removed my profile, closed the Relative Finder “feature”, and removed all my sharers. The 23andMe site is shite.
I went to 23andMe to find out my ancient genetic history, and my raw SNP data. I got my raw SNP data minus some no calls which the company does nothing to find out why there are no calls from their chip. As SusanC indicated the health part is of little use, and as sound as a bull market at full charge. It panders to the hypochondriacs of which are are many at 23andMe. Ancestry, forget that one. They tell you nothing you don’t know already. They have enticed a lot of adoptees or children and grandchildren of adoptees with false hopes that the company can assist them in locating their parents, grandparents and so on. They haven’t helped one person find out any more than they can by going through adoption records, census forms and other genealogical methods. SusanC also indicated the uselessness of the RF feature. You find out you are related to anonymous strangers in far away countries with indicators of totally different geographical origins around this planet. All the hard work of finding a provable connection with these remote and obscure relatives is left to you. Gee thanks 23andMe.
23andMe’s product is shoddy compared with deCODEme, and the health things can be gotten from the Promethease Report. Dr Doug McDonald’s use of my raw SNPs gave me a realistic idea of where I fit into the genetic landscape of Europe, and with better PCA maps. I have wandered over to deCODEme. The company has a reasonable clientele, less hypochondriac Jewish Princesses and badly educated ignorant American citizens. That brings me to another problem. 23andMe’s clientele is overwhelmingly North American mixed North and Central Europeans, and Ashkenazim Jews. It is futile for me, a Southern European from Europe and unmixed, to compare myself with North American mixed Northern/Central Europeans, and American Ashkenazim Jews. deCODEme has less of the mongrels so to speak and a better representation, and less mixed ethnic European nationalities and other groups from the Middle East, South Asia and East Asia.
Don’t go to 23andMe. Save your pennies and go to deCODEme.
“The 23andMe site is shite.”
No, it’s not. The 23andMe site is attractive, functional, easy to navigate, and reliable.
“the health part is of little use”
No, it’s not. The 23andMe health reports already include the same tests ordered by medical doctors to make medical diagnosis. For example, Aetna considers the BRCA1 and BRCA2 founder mutations tested by 23andMe to be sufficient to justify coverage of a prophylactic mastectomy. In other words: your 23andMe test is enough to get medical insurance to pay to cut off your breasts without yet having breast cancer.
http://www.aetna.com/cpb/medical/data/200_299/0227.html
I have picked this specific example because it is bombastic and not ambiguous.
Less critically but more widely applicable: drug metabolism can be predicted using the 23andMe test. For example, the FDA has added a boxed warning to the common blood thinning drug clopidogrel for mutations in CYP2C19 tested by 23andMe. In other words: your 23andMe test is enough for the FDA to announce that a common drug is not effective in people with the CYP2C19 mutation.
http://online.wsj.com/article/SB10001424052748704131404575117761605984580.html?mod=WSJ_latestheadlines
Listen idiots: you got used. 23andMe is not a shoddy a product. 23andMe is an excellent product. 23andMe is rapidly becoming a medical revolution. The problem isn’t the product. The problem THE COMPANY. 23andME FUCKED UP. They don’t care about getting medicine to people. They care about staring a fucking “healthquake” on Twitter. You know how I know? THEY SAY SO ON THEIR FUCKING WEBSITE.
LET IT DIE.
There are plenty of other genomics companies. But they won’t get a chance to do anything good if the 23andMe zombie sucks out the sunlight or _more_ completely fucks up legal and market precedent with their shambling antics across the cocktail lounge.
But hey, maybe you know something that the Brins don’t about genomics. I’d love to hear the defense for 23andMe when even its founders don’t trust to keep their money it.
LET IT DIE.
Genomics will be better for it.
Edit: I got used. Maybe you didn’t.
But maybe I never got the memo on “How to use a 23andMe test in medical practice if you actually own both a medical practice and a 23andMe even though you are contractually bound by 23andMe not to use 23andMe in medical practice.”
Maybe that part was on Oprah. I missed that episode.
Fuck, I done did me a Steven Murphy on this here comments thread.