Monday, March 24, 2014

Startup advice - love your co-founder!


Miguel Hernandez, founder of Grumo Media, just interviewed me for a cool new series that he started "Startup Stories" (Grumo Media is the agency that made our amazing PubChase cartoon).

I answered many questions and shared a lot of advice for entrepreneurs from my two years of working on ZappyLab, but my most important tip from this interview is on the topic of picking a co-founder for the startup.  Many have blogged on this topic because the choice of the co-founder often makes or breaks a fledgling company. My personal favorite is this post from Naval Ravikant, the co-founder of AngelList.

Naval, like many others compares the choice of a co-founder to the choice of a spouse. I agree entirely, but I think it is important to realize that the relationship with the co-founder is going to be tested and strained even more than a marriage.

Over the past year, I have spent way more time with Alexei, the co-founder of ZappyLab, than with my wife. And our time together is not a hike or picnic or a romantic stroll on the streets of Paris. We are in a basement. We are stressed. We are sleep-deprived. We are anxious about funding. We are constantly making decisions and prioritizing. Some of the decisions, if made incorrectly, can kill the startup. We debate and disagree (that’s the best part of having a co-founder – the ability to disagree and debate; without it, your ideas won’t improve and your startup will fail). And the worst thing is that Alexei and I don’t have sex with each other. So unlike a marriage, there is more stress, more arguments, and no cuddling and romance to dissipate the tension. You better have the right co-founder for this.

(I have been married for 15 years and in the startup mode for 2.)
--------------------------------
Here is the entire interview, and the rest of the answers are pasted from Miguel's post, by permission, below.

What is PubChase?
It is a personalized recommendation service for biomedical literature. A bit like Netflix predicting which movies you might like based on those you’ve seen, PubChase recommends new research papers based on the ones you have in your library.

How did you come up with your idea?
I didn’t. Matt Davis did. He is a scientist extraordinaire, currently doing postdoctoral research in Heidelberg, Germany. Matt came up with the original PubChase recommendations idea. He cares deeply not just about the questions in biology, but about the way science is done, openness, and the scientists doing the research. He wanted to make scientists’ lives easier. That is why he has not slept for the last two years, while finishing his graduate work, starting a postdoc and simultaneously working on PubChase.

I came up with an entirely different idea in January 2012.

We hear about groundbreaking discoveries and life-changing new technologies, but the reality is that most of the important discoveries are never published or shared at all.
Edison had to reinvent the light bulb, filling 40,000 pages of lab notes and laboriously testing more than 1600 materials for the right filament. Had all the scientists working on it been able to connect in real time and share their discoveries, there is no doubt that the light bulb would have evolved substantially faster and more efficiently.

The same thing that was true in the 1800s is still true today. Scientists are constantly re-discovering knowledge that others have not had the time to publish and improving existing methods without the ability to share the improvements with the world. Imagine what the pace of Scientific progress could be with the aid of modern technology. This is why I co-founded ZappyLab, so that the next great discoveries in Science will come 2, 10, 35 years earlier than they would have otherwise.

I imagine that our project (protocols.io) may appear to be very targeted at scientists rather than the general public. But our enterprise isn't just an attempt to create a useful resource for scientists; this will revolutionize research in the lab, speeding up discovery and saving society billions on wasted efforts to reinvent the wheel. Vaccines, climate change, aging, cancer, and basic research the work that impacts everyone and everything will move significantly faster because of what we are doing. This isn't a project for scientists. It's a project for humanity and our world.

So why did you build PubChase?
We just launched protocols.io – a knowledge sharing platform to instantly communicate corrections and optimizations to science methods.

The challenge is that this resource has to be crowdsourced, but you cannot crowdsource without a crowd. So, for two years now, we have been bringing mobile technology to laboratories (phones and tablets for now and Google Glass in progress), in order to create a knowledge-sharing infrastructure that will fundamentally transform the communication of discoveries between researchers. We built PubChase also as part of this to create a community and acquire users towards the protocols.io effort.

How did you validate your idea?
Beta testers! Talk to your users. Get feedback. Refine, improve. Get lots of users to be your consultants, exclusive club members, whatever you come up with – but get the users to give you feedback.

Why did you become an entrepreneur?
Accidentally. Wanted to be a professor. But Alexei, co-founder of ZappyLab, called and asked if we can build a mobile app for biologists. I first made fun of him, but then came up with an app that would be useful to me. Three weeks later, realized that this app could help to create protocols.io. And then, ZappyLab, little by little sucked me in. I became obsessed with the protocols.io idea. Given its potential to have an impact on this world, I realized that if successful, I would contribute so much more to society through this than through any discovery I would make in my own lab.

For a year, I continued research by day and ZappyLab by night (even after we raised seed funding). It was still, somehow, a very productive year for me scientifically. But then I realized that I couldn’t start my own lab, and make sure that ZappyLab succeeds. I could be certain to fail at two things, or decide to try to succeed at one.

How did you fund your startup?
Borrowed, savings, loans, credit card debt, angel investments, Kickstarter – everything! (Initially, I borrowed $50K from my parents and Alexei and Irina invested $50K from their savings. Then angel seed round for $325K.)

Share 1 tough moment/challenge in your journey
No such thing as a tough moment. It’s a roller coaster. It’s been two years of tough moments and challenges. That’s the startup life. But it’s also exhilarating and incredibly rewarding (just not financially).

For me and Alexei, probably the hardest and scariest is the initial decision. Alexei had to give up being a CTO of another really successful startup. Give up a really good salary and stability. I had to give up on the dream of becoming a professor and teaching and doing research. This initial decision is the hardest because you have a choice. It’s scary. It’s an enormous risk, and we realized just how great the risk we were taking. However, once you make the decision and plunge in, there is no way back. You have no choice but to solve all the other problems. This is also why you have to commit 100%. You can’t do it on the side. You can’t have a side-startup. Even if you plunge in, your startup is likely to fail. If you do it on the side, it’s guaranteed to fail.

Share 1 breakthrough that helped your startup
The right team in terms of the co-founders. And I don’t mean just good experts. Sure, I spent over a decade as a scientist – I know what research, publishing, grant-writing, and discoveries are like. I know what to build. Alexei has spent just as long building web and mobile tools of incredible complexity and beauty. And he is an absolute super-human as a team leader and engineering and technology guru.

But there are thousands of scientists like me with the same ideas that I have. And there are probably a few hundred software engineers who are technically as good as Alexei. That doesn’t mean there are hundreds of startups like ours.

By the right team I mean – luck to know each other. And luck that we are as close of friends as we are. No matter how technically great the co-founders, there will be tough moments. No, let’s be more clearly – there will be one long tough moment for the entire duration of the startup. And you need the right partner to pull this off.

Many people compare the co-founder relationship to a marriage. I think it’s harder than a marriage. I have spent way more time with Alexei over the past year than with my wife. And our time together is not a hike or picnic or a romantic stroll on the streets of Paris. We are in a basement. We are stressed. We are sleep-deprived. We are anxious about funding. We are constantly making decisions and prioritizing. Some of the decisions, if made incorrectly, can kill the startup. We debate and disagree (that’s the best part of having a co-founder – the ability to disagree and debate; without it, your ideas won’t improve and your startup will fail). And the worst thing is that Alexei and I don’t have sex with each other. So unlike a marriage, there is more stress, more arguments, and no cuddling and romance to dissipate the tension. You better have the right co-founder for this.

Share 1 piece of advice to upcoming entrepreneurs
I'd like to share 2 pieces of advice:

a. Get experienced advisors. Either through an incubator or by networking. There are many people who have started companies. They have a lot of wisdom. A lot of them love mentoring and helping talented and passionate new founders. Seek their help. Get lots of advice.

b. Some times, you will get terrible advice, even from the people who are more experienced and are trying to help you. Know when to ignore the advisors from (a). Trust your instinct. Pretty quickly, you will become the expert. You’ll know your product, users, and vision better than anyone in this world. Trust your intuition.

Lenny's recommendations to other entrepreneurs


Share 1 entrepreneurial blogs
Fred Wilson’s AVC.com and Paul Graham’s paulgraham.com are treasures.
Share 3 platforms/tools you cannot live without.
AWS, Google (gmail/analytics/Docs)
What do you use for project management? (Basecamp, Asana)
Used JIRA and Basecamp. Now all on GoogleDocs and Wunderlist.
How do you do your accounting? (Quickbooks, Wave, Bookkeeper)
Quickbooks.
Share 1 Productivity hack.
Amazingly talented engineers in Moscow, RF.

About the demo video Grumo Media produced for PubChase


How did you find Grumo?
Via the Grumo Quartzy video.

What made you decide you needed a demo video?
Once your product is complicated enough, you realize you need an FAQ. You write an FAQ. Then you look at GoogleAnalytics and realize you are the only one who has been to the FAQ page in a month, and none of your website visitors are registering. You think about that for a minute and then ask yourself, “Did I ever look for an FAQ on a new website?” And that’s how you realize you need a demo video.

What have you done to promote your video?
Great question! We built an amazing product (PubChase) and it turns out no one knows it exists. Need to market it! A demo video seems like a good idea. But once you put it up on youtube and your landing page, it hits you that no one knows you have this amazing demo video from Grumo.

We ran a fun contest where we rewarded with a t-shirt those would could identify the Nobel Prize winner in our cartoon. Ultimately, unless you can magically get the video in front of the right, it does not necessarily market for you. But if you do the marketing right, the video will help greatly with conversion – convincing the visitors to your website to register and log in.

Wednesday, March 5, 2014

My Life as a PhD Scientist – You Should Know Why Science Will Fail

[Anonymous essay, e-mailed to me in response to my "Goodbye Academia" post.]
 

I am the “best and brightest” humanity has to offer.  I am in the top 1% academically.  I obtained a PhD in Genetics from an Ivy League university.  I became a postdoc at one of the most prestigious universities in the world: the University of Oxford in the UK.  For the last eight years I dedicated my life to cancer research, making great strides for Personalized Medicine, work that saves lives.  And what do I have to show for all the long hours, the high-stress, and the good I do for society?  Nothing.

I am broke.  The System is broke.  And this is my story.

I grew up in rural Tennessee.  I always dreamed big.  I wanted to do something important with my life, help people, and really live it. The best thing for me as a child was not knowing how difficult it is in the southeast to have upward mobility. 

I was the lucky kid who never had to study for tests.  I always scored in the 99% percentile on the annual state assessments.  The school administration tried to skip me a few grades along the way, but my mother was adamantly against it.  When I went to college, I had a Presidential Scholarship, a free ride, because of my academic excellence.

As an undergraduate, biology drew me in, particularly the cool science of genetics.  It was an amazing world where we were just unraveling the human genome, figuring out what bits of our DNA did what.  I realized that if I pursued my PhD in genetics, I could work on cures for diseases and potentially save more lives than even the best MD. 

Through perseverance and hard work I made it into Cornell University. I specialized in cancer genetics, and my graduate work culminated in a discovery that will help guide treatment decisions for the 283,000 breast cancer patients annually whose tumors have lost a gene called NF1 (>25% of the 1+ million annual breast cancer cases).

But there was trouble in grad school that had nothing to do with hours, stress, or troubleshooting experiments.  Money.

Cost of living is very expensive in NY.  My stipend barely covered enough for me to break even each month.  I had to be frugal with my money, seldom going out to movies or dinner.  My parents always had to fly me home for Christmas.  I told myself it was only temporary, that I’d make more money once I became a postdoc.  It was a long six years.

My money wasn’t the only money problem.  There was no money for the research.  My PI (the professor and head of the lab) got rejected time after time for NIH grant funding (and every other funding organization out there).  Ivy League, and he was still relying on his start-up money to cover the costs.  The stats at the time (which haven’t improved) were that < 3% of all grant applications were funded.  Unfortunately my boss was never part of that 3%.  In the six years I worked there, he never got a single grant for the cancer research.

I worried as graduation drew close.  I had limited time to find my next job.  And even if I did, would I ever actually be able to get a faculty position?  All of these labs are run by an army of students and postdocs, but there’s only one guy at the top, one PI for each. 

My boss was on the hiring committee for the department.  He told us the inside horror stories of what goes on behind those closed doors.  The national average at the time was that for every one faculty position, there were 200 applications.  For our department, there were 300 applications for every one faculty position.  As if the odds alone weren’t daunting enough, he told us how the hiring committee got through so many applications.


  1. University the applicant came from.  If the applicant wasn’t Ivy or from one of the best Tier 1s like Harvard, Stanford, or MIT, the application was immediately thrown in the reject pile without a further look.
  2. Top publication record.  If the applicant didn’t have publications in the top journals like Science or Nature…reject pile.
  3. Name of the lab they came from.  If the applicant’s boss wasn’t a Somebody…neither was the applicant.  Reject pile.
  4. Contents of the CV / Resume.  Finally, at step 4, the hiring committee would actually bother to look at the real contents of the CV / Resume, like how many publications the applicant had and teaching experience.  None of that meant anything if the three previously mentioned items couldn’t save the applicant from the auto-reject pile.

The stats said that only 1/3 of PhDs would even stay in science because there were so few jobs (academia, science writing, industry, etc).  That’s right, after 6 years of grad school and even 10 years of postdoc, 2/3 end up outside of science altogether.  I knew several myself.  One girl finished her Ivy League postdoc and took a job in retail.  A guy I knew quit and became a stable-hand at a horse stable.  Another went to work at a bakery in Boston.  Another traveled the world for a while doing nothing – never knew what became of him.  The world’s “best and brightest”, their talents and skills completely forsaken and discarded by the system.  It’s not just them who suffers.  It’s every human being who will never benefit from the discoveries they could have gone on to make.

But maybe they were the lucky ones.  They got out before they lost any more of their lives.  For those scientists that made it all the way into a faculty position, there was still a high chance that they wouldn’t make tenure.  New faculty get five years to get funding and churn out publications.  Otherwise, sayonara.  They get a year to find another job (yeah right, who will take them after they failed to get tenure at their current job), and then they have to pack up and leave.  I knew a female PI who started around the same time I started grad school.  She went on maternity leave along the way.  Having children was important to her.  She wanted a family.  It hurt her publication record and funding chances.  She didn’t get tenure.  She was gone before I graduated.

Douglas Prasher got the shortest end of the stick.  He should have gotten the Nobel Prize.  He got the opposite.  Prasher cloned and sequenced the gene for Green Fluorescent Protein (GFP).  And Prasher was the first to propose that GFP could be used as a tracer molecule.  He wrote a grant detailing how GFP could be used as a reporter to measure the levels of gene expression and track the localization of proteins in cells.  Every geneticist and molecular biologist alive today uses GFP assays in their research.  GFP revolutionized the field and allowed scientists to make rapid leaps forward.  Not surprisingly, the work for GFP received the Nobel Prize.  Only, Prasher didn’t win it.

You see, that grant he wrote about GFP never got funded.  The reviewers of the grant thought his ideas were crazy.  Without funding, Prasher didn’t get tenure.  He had to close his lab.  He gave his GFP samples and ideas to his fellow colleagues.  He knew how important GFP was, and he didn’t want it to be lost.  With no funding, Prasher was forced out of science altogether.  He became a shuttle bus driver for a car dealership. 

The semester I taught the Principles of Cell Biology and Development lab course at Cornell, I made my students write a report on Douglas Prasher.  I wanted my students to know what they were getting into.
Graduation finally came for me! I proudly marched my parents around the campus after the ceremony.  I was doctor!  My work in cancer would help a lot of people!

But that didn’t seem to help me get a job.  Knowing the potential dead-end of a postdoc position, I first looked into pretty much every other area of science.  Science writing, industry, consulting, and teaching.  Nothing.  So, I caved and hunted for postdoc positions.  I sent out over 150 queries and applications.

In the end, my tenacity paid off.  I had a choice between three excellent universities: Harvard, Cambridge, and Oxford.  I found from the lab members during my Harvard interview that Harvard didn’t pay its postdocs’ health insurance (at least not in that department), but Harvard required the postdocs to have it.  That’s right.  No health insurance, even for the full-time “best and brightest”.  And Boston was as expensive as NY.  The postdocs only made financial ends meet by relying on spouses or roommates.  Between that and the 80 hour work weeks the lab members warned me about, I turned Harvard down.

Cambridge and Oxford claimed to pay better than their US counterparts, and at face value, it certainly appeared that way by looking at the salary conversion from pounds to US dollars.  I loved both places, both projects seemed fantastic, and the PIs seemed great.  If nothing else, the UK had a great work-life balance that academia in the US completely lacked.  Many of the labs worked standard 40 hour weeks (contractually it even states 37.5 hours). 
When I chose Oxford, I was full of excitement and hope!  A new country!  45 minutes to London!  An hour and a half after that to Paris!  The city is beautiful, full of ancient architecture that has survived thousands of years.  Home to the Bodleian! 

I love my job at Oxford.  The cancer project is super interesting with a focus on discovering Biomarkers for use in the clinic to improve patient survival and outcome.  My boss is great, and I work normal hours.
But.

Money.

Money is a major problem.  I don’t have a house.  On a single income, 65% of my salary is burned directly on rent and Council Tax.  By the time food and utilities (phone, internet, electricity/heat, water) are added in, I make $0 at the end of every month.  And because I make $0, I can’t save anything to turn things around, like a down payment for a house.  Now, I don’t go to the movies at all.  I never go out to eat.  I’ve never been to London, much less Paris.  I can’t afford a pet deposit to get a dog.  I don’t have a car.  I don’t have cable.  I don’t have a TV.  I don’t even have living room furniture.  It’s an empty room with a modem and the rug the previous owner left behind. 

It’s just a phase.  That’s what I’ve been telling myself since 2006.  And maybe it’s not a phase.  I really want to teach, but I know the odds of landing a faculty position - 200:1 anywhere I go (300:1 at a Tier 1 institution).

And every scientist I know is in the same predicament.  Science will fail because the System is running the scientists out of it.  Every human suffers because of it.  Cancer affects 1 in 3.  Someone close to you will get it.  How many could be saved if the system didn’t fail scientists like Douglas Prasher?  Like the 2/3 of the PhDs that are forced out of science?   

I am one of the “best and brightest”. Ever since I was a child I felt special, like I was meant to do something incredible with my life.  And I have.  Only, being a scientist has come at the complete sacrifice of my own life.  I put the last eight years of my life on hold, and I’ve set myself up to keep it on hold for the next 5-10 years as a postdoc.  But I’m tired of this.  I want to go to the movies.  I want a dog.  I want a family.  I want to live my life.

Sunday, March 2, 2014

Hello Startup (the sequel to “Goodbye Academia”)

My older daughter Eeva was born on January 8, 2010. Two years later, on January 7, 2012, ZappyLab was born. I was buying food at the Cambridge Trader Joe’s for Eeva’s birthday party. Alexei called me while I was in the store and asked if we could build an app for biologists. Irina and Alexei (the co-founders of ZappyLab) were making kiddie apps on iOS for a while, hence this question. I told Alexei that phones have no place in the lab, and suggested that instead he build an app for African safaris, so that when someone sees a hippo, they can share the coordinates and have other people come to the spot (yes, I have a million stupid ideas – that’s partly why every startup needs a partner – to weed out the stupid ones). Twenty minutes later, I thought of a protocol checklist app. That’s something that would be useful to me in the lab.

I called Alexei, he said Irina would build it with his help; they know apps and I know science and scientists. We would give scientists a terrific productivity tool and it would spread like fire quickly with a million downloads at $1 each, because no one else has made anything like this. The night after Eeva’s wonderful birthday party, I stayed up and wrote the specs for the protocol app. The specs included the essential ability to track changes in each run of the protocol and record in the app precisely what was done in the experiment that day.

Three weeks later on January 27, walking home from lab, I called Alexei and asked if we could add a “share” button to the app and crowdsource a central protocol repository. For Alexei, everything technical is easy – he has been doing exactly things like this for ten years. The only wrinkle in our plan – if we want to build a crowdsourced repository, the app has to be free, and of course so does the repository. The revenue from the app sales suddenly evaporated.

I called Michael Eisen, described the idea, and asked if PLOS could build it. Mike said that the idea is genius, but we have to make it happen, not PLOS. I called Jasper Rine and asked how to do it, without quitting the postdoc; a nonprofit? Jasper said that there is no way I would get a grant for this as a postdoc and made it clear that what I wanted to do is nearly impossible without quitting.

We needed a plan B. At 3am on Saturday, a week later, I woke up and figured out how to do it. What do all scientists do? Count! We would build the Lab Counter app (now inside Bench Tools), sell a million copies of that, and use the proceeds to build the protocol app and repository. I quietly crawled out of bed into the bathroom (parents were visiting and sleeping in the living room, and I could not write anywhere else). For the next two hours, I sat in the bathroom writing the specs on for the Lab Counter. Called Alexei and Irina in the morning and we decided to build this. Irina would take the lead on developing this, while also taking care of three little daughters. Alexei would help her, while working as a CTO of a fashion-blogging startup. (Yes, even with a partner or two, you will have naïve ideas that all three of you buy into.)

Took us a month after that to understand that we needed another plan. I had no plan any more, but I knew we had to build the protocol repository. I knew we had to try, or I would never forgive myself. So I picked up the phone, and called my ex-boss Joe Duncan. Asked him if he would be an angel investor and help to make this happen. Next day, after leaving the lab, instead of going home, I took a train to New Jersey and spent the weekend telling Joe about the idea. He got it instantly. And he agreed to be our lead angel investor.

Two days later, Alexei registered protocols.io. Doing the paperwork and getting the angel investment would take too long. We knew we had to start and would do it. So I borrowed $50K from parents. Alexei and Irina put $50K of their money into the company. We incorporated, Alexei resigned as the CTO of the other startup, and for almost two years, he has been in the basement, working insane hours every single day, to build everything that we have built.

I’ll skip most of the two years and fast forward to today. We understood that there is no crowdsourcing without the crowd, and have done everything possible to give scientists amazing free tools and resources to build up our user base. And we have built the protocol repository and the checklist app. But getting users is much harder and slower than anyone ever imagines. And almost nothing goes viral on its own. That’s why I had to leave my postdoc.

But funding ZappyLab is hard, when everything you are building is free (indeed, research or startup – funding is never easy). We will be sustainable because we will charge vendors for links to their chemicals. But we can’t do that until we have hundreds of thousands of users. We are firmly a pre-revenue company.

Both of our families have burned through all of our savings. We are deep in debt. Maxed-out credit cards. Four weeks of funding left in the ZappyLab bank account. We have given everything we have financially, emotionally, and physically to the idea of the protocol repository. We have nothing left to give.


That’s why the Kickstarter campaign – you scientists have to help us make this happen. Or it won’t. And if we go bankrupt, it is possible that someone else will create it. But consider that no one has managed to pull this off until now. What if without us it takes another 20 years? What if it never happens? This is something that simply must exist. You may save yourself a year of work by backing us on Kickstarter with a few bucks. You can help to save billions of dollars for our society and to speed up research. It’s worth it.