It was a treat to host Fred Wilson of Union Square Ventures at Harvard Business School today – his first time attending a class at the school.  Fred, as most readers of this blog know, is a venture capital legend in the making and the investor in some of today's leading consumer Web properties, including Twitter, Zynga and Four Square [Fred's post on his visit can be found here].

Fred and I had a discussion about lean start-ups and pattern recognition with the HBS students in Professor Tom Eisenmann's class "Launching Technology Ventures".  If you want to see some of the Tweets that came out of the class (imagine a professor encouraging students to grab their smart phones and live Tweet in class!) you can check them out here (#hbsltv was the hashtag).

A few takeaways from our session that I thought were particularly insightful:

• Early on in a start-up, entrepreneurs should be hunch-driven more than data-driven.  If you are only data-driven, the risk is that you will move too slowly.  It's more important to have a hypothesis about what might work and what might not work and then see what happens in the marketplace to prove or disprove that hypothesis.
• Lean start-up as a methodology or approach is very useful, but isn't a guarantee for success by any stretch.  Think of the methodology as a machine.  If you have garbage inputs, you will still have garbage outputs.  There's no substitute for good strategy, great entrepreneurs and a very large market opportunity.
• When considering when to monetize your new product/service, think carefully about whether the monetization strategy actually improves the service or is a distraction.  Banner ads on Facebook are a distraction (as Zuckerburg supposedly said in the movie Social Network, "No ads. Ads aren't cool.")  But, for example, on Etsy if someone pays for a product, it inspires producers to create more products.  Thus, the monetization is harmonious with building the service.
• If you are going to fail, and certainly with more start-ups being created and seeded we will see more failure, be sure to fail gracefully.  How you handle yourself as you unwind / seek a soft landing will reflect heavily on you and will cement your reputation.
• Don't worry about whether you are building a feature, a product or a company.  Build something great, have huge passion for it, engender affection with a large customer base, and let the rest follow.
• If you get traction, transform your company into a platform.  The most valuable companies are those where third parties help you grow by plugging into your services like a utility.
• VCs don't make companies successful.  They can believe in and support a company, but ultimately the entrepreneurs make or break the company's success and don't let anyone (particularly an egotistical VC!) imply otherwise.

As we ended the class, we tried to inspire the students to "go for it" and become entrepreneurial.  I am always pushing students to consider if now is the right time for them (see my recent blog post:  "Should I become an entrepreneur?") and pointed out that this was a time in their lives where they could afford taking more risk.  Once they get married, have kids, buy a house and get a mortgage, it's a different ballgame.  Fred quoted a friend who once told him there were three addictions in life:  "calories, heroine and a paycheck".  If you can break the last addiction, you are well positioned to become a potent entrepreneur!

The Wall Street Journal's article today about the existence of "alcoholism genes" and what the future might bring ("imagine you go to your doctor and say 'I'm drinking I need help.' and they do a blood test and, if you qualify [based on genetic markers], they give you medicine the next day.") is a part of a larger trend that will radically change the world's health care system.  With a nod to my friends at Hubspot, I'll refer to this future phenomenon as "inbound marketing comes to health care".

First, some background.  The cost of mapping your genes is falling rapidly.  Today, you can get your genes mapped and analyzed for $10k.  In 3-5 years, that price will fall to $1k.  Harvard Medical School Professor George Church, one of the great pioneers in this field, observed recently to one of my partners that, "people will spend $10k per year on insurance but over a 70+ year lifespan are not yet comfortable spending $10k for their genome to be sequenced." As the process gets cheaper and the data and analytics gets better, that will change.  I'd be shocked if you, dear reader, did not have an analytical report in your files somewhere in 5 years about your personal genome and insight into its health implications.

And that gets me to the concept of inbound marketing.  Inbound Marketing (as captured nicely in the book by Dharmesh Shah and Brian Halligan), is the notion that the new era of marketing is about pull, not push.  Rather than producers pushing their products onto consumers, consumers have the tools and means to show up at the producer's door "inbound" and identify their needs and interests.  

It's become very clear how this technique applies to products – consumers research their needs by searching this huge information database in the cloud called "Google" and find what products and services might serve their needs and proactively contact and, eventually, purchase those products.  This technique is why many companies invest so much money in search engine optimization (SEO) and search engine marketing (SEM) – a business that has grown to tens of billions of dollars and fueled Google's meteoric rise as one of the most successful companies and global brands in business history.  Businesses are redirecting their tens of billions of "push" marketing dollars  into other mechanisms that set themselves up to be found by intelligent, informed consumers.

Now, let's go back to health care.  Imagine that in 5-10 years that tens or even hundreds of millions of people have their genomic data stored in the cloud.  Imagine that this data can be indexed, analyzed, parsed, sliced and diced.  And imagine that it is very, very secure.

What might happen with that kind of large-scale genomic data available in that format?  Inbound marketing.  Rather than pharmaceutical companies pushing drugs through their large sales force, they can access this database and alert consumers as to what drugs might fit what genomic profile.  Rather than hunt for clinical trial candidates in hospitals throughout the world, drug companies can email the relevant 1000 patients that precisely fit the indication they would like to test. 

Let's make this very personal.  My father-in-law recently died of ALS.  His older brother also died of ALS a number of years ago.  Thus, there is a reasonable chance that my wife's family has some genetic predisposition to ALS.  In today's health care environment, where genomic information is expensive and sitting in silos, there is nothing much we can do about it but wait and worry.  But someday in the future, perhaps as soon as 10 years from now, we will have the opportunity to opt-in to a service that will alert us via email or text when ALS drugs that might address this particular issue enter clinical trials.  Or perhaps even approved by the FDA.  We might all register our genomic data into this service so that we can receive alerts and information about any range of insights or treatments that might be relevant to our personal make-up.  This is "personalized medicine" in the extreme.

One of our portfolio companies, Predictive BioSciences, is pioneering a urine biomarker technique – pee in a cup, and Predictive will tell you if you have cancer.  In the future, we might all be swabbing our cheeks, peeing in cups, and pricking our fingers to tell us much, much more.  And when that information is available to our trillion-dollar health care infrastructure, imagine the possibilities.