New blog at a brand new place!

Hey everyone,

I decided it was about time to redesign my site and get a “real” domain.  Future posts will now go up at, and  my newest post is up on that site today. All of my old posts have also been ported over.  I hope you all will continue following me at!




Online Education: How to Engage Students?

Massively Open Online Courses (or MOOCs) are classes taught online that integrate “traditional” course materials like recorded lectures and problem sets with tools that can only be used online, such as interactive forums or live chat windows. Sources from The New York Times to The Economist have touted MOOCs as “the future of education,” arguing that the Internet and online classrooms provide a medium that, in comparison to the classic in-person lecture model, is incredibly well suited for learning material. Online education startups like Coursera, edX, and Udacity are free to use, and they’ve enabled millions of students from around the world to study subjects they would never have been exposed to otherwise.

One major concern with MOOCs, however, is around user engagement. A University of Pennsylvania study found that course completion rates are incredibly low, averaging out at 4%, and the only course characteristic proven to boost completion rates is a lower workload. While students learning online may perform better than those who study in a traditional classroom setting, students struggle to commit to online courses long-term. How can we make online courses more engaging?

Argument for “Embodiment”

Natasha Myers, a professor of Science and Technology Studies at York University, argues that in order for students in Biology to deeply engage with course material they must physically interact with what they’re studying. Over the past 30 years, scientists have grown increasingly reliant upon digital three-dimensional models that capture the nuances of proteins, a microscopic physical object. These digital models are very effective in capturing the nuances of a protein, and Myers recognizes the temptation in solely using this most accurate representation in the classroom. By converging these digital representations with something that students can interact with in the real world, however, Myers finds that in order to generate a “complex mental image” that would be absent otherwise.

For Professor Jim Brady, one of the academics Myers studied, this means literally making his physical body mime the behaviors of a protein, and getting his students out of their chairs to follow suit. In his classroom, Brady has students stand up and cross their arms to demonstrate the packing of helices during folding, and to explain how collagen fibers form in wounds he has students examine their own scars. By literally making students more personally connected to what they are studying, Brady finds that his students become more engaged with their course material.

Is “Embodiment” Necessary?

It is tempting to extrapolate the results of a study like this and say that online education fails without an in-person component. Many, including the higher-ups at my university, are pushing for a “hybrid” approach that links “passive” lectures given online with an in-person “active” learning component that emphasizes personal engagement with the ideas discussed in those lectures. This approach has negatives, however, in that it goes against two of the primary appeals of online education by limiting class sizes and making courses far more expensive to run.

It is important to note that Myers is making her claims about embodiment for one distinct field (Biology). While embodiment is useful in helping students understand microscopic physical phenomena, it is probably less useful in helping students learn Computer Science. It makes a lot of sense to learn about how computers work while using a computer. Ultimately, the way that a course is taught should fit in with the material being discussed. For some courses a physical relationship to the material being discussed may be important, but others may thrive without this added constraint. There is no “one size fits all” solution for increasing engagement in online education, but there are general principles that should be considered.

Making Online Education More Personal

The incompletion rate for online education courses is high, not because they lack a real-world component, but because people do not see the real-world, personal implications of online studying. When the course becomes difficult or the workload too great, it is easy to close down the browser and never sign back when you feel like the course has no bearing on your day-to-day life. One online education website that has thrived off of creating a real world connection is, which donates ten grains of rice to someone in need for every correct answer to a vocabulary question. Over 100 billion grains of rice have been donated through Freerice since its inception. Not every course can make its material seem real in such a literal manner, but every course should strive to make its material seem similarly important.

Why America Needs Companies Like Twitter

Nearly every day news publications highlight recent graduates who are “choosing Silicon Valley over Wall Street”, embracing not only the startup culture and lifestyle but the dream of getting rich quicker.  It is therefore understandable that Twitter’s initial public offering was a major event not only on Wall Street but college campuses across the world.  Twitter follows a long line of consumer-facing tech giants to price their IPOs in the last five years, including Facebook, LinkedIn, Groupon, and Zynga. But according to Dan Primack, a prominent tech blogger for Fortune magazine, there doesn’t seem to be another similar company on the IPO horizon any time in the near future, at least not one on the scale of “even my mother would know about it”.  While Primack notes that it seems like this should only matter to venture capitalists and bankers, the lack of prominent consumer technology IPOs could have a larger effect that permeates not only the entrepreneurial ecosystem, but American society as a whole. 

The phrase “technological sublime” was first developed by Perry Miller in his study “The Life of the Mind in America”.  A scholar of early American history, Perry wrote that “technological majesty join[ed] with the starry heavens above and the moral law within to form a peculiarly American trinity of the Sublime.” The steamboat, for example, evoked a sense of wonder typically reserved for God. The steamboat became “a subject of ecstasy for its sheer majesty and might, especially for its stately progress at night, blazing with light through the swamps and forests of Nature,” inspiring Mark Twain in the same way the stars inspired Van Gough.

David Nye expanded on this idea in his book “American Technological Sublime”.  Using case studies ranging from the skyscraper to the atomic bomb, Nye demonstrated how the American response to certain technologies is best described through the lens of the sublime.  The phenomenon of the technological sublime is as old as America itself. In an earlier era, Americans thronged at the 1939 World’s Fair; today we wait in long lines at the Apple Store on release day.  The Brooklyn Bridge and the Hoover Dam have been replaced by the iPod and the iPhone, but the same veneration of technology remains, giving a common source of devotion to a nation devoid of cultural and religious homogeneity.

Perhaps the most telling example of the technological sublime can be found in The Social Network.  David Fincher’s movie recounts the early story of Facebook through the lens of its founder, a self-assured college dropout who changed the world from a Harvard dorm room.  The movie went on to receive eight Academy Award nominations, including best picture, and prominent film critic Peter Travers of Rolling Stone named The Social Network his movie of the year.  Mark Zuckerberg’s story evokes a sense of awe among my peers, and it’s not because of the effectiveness of his customer acquisition strategy or the way he structured his data centers. Facebook became my generation’s moon landing, arousing not only admiration but national pride.

Will this same fervor greet the tech successes of the future?  Veeva Systems provided a better return on capital for its investors than Twitter, but none of my friends shared that story because Veeva is an enterprise company that provides content management solutions for the life sciences industry (and that’s not even fun to read).  Zulily went public at a $2.6 billion valuation, but barely a single tech blogger wrote about it because it has a functional revenue model and its co-founders are in their 40s. Sexy companies are started by college dropouts and don’t make any money.  Those companies evoke the sublime.  Real businesses with uninteresting founders belong on an investor’s Excel sheet, not on the front page.

While companies like Uber or Spotify could rise to become the next big consumer success stories, chances are we will have to wait at least a year and probably even longer for our next big consumer tech IPO.  This leaves the question: what happens when the technological sublime runs dry? As Michael Sacasas notes, throughout history Americans have been divided on the basis of class, ethnicity, and religion.  Americans share no strong historical blood lines or ancient ties to the land.  What has bonded Americans together, however, is a belief in technological progress reinvigorated by one sublime technology after another, from the transcontinental railroad through The Social Network. I worry that the coming dry spell will impact not only Silicon Valley and Wall Street, but our sense of what makes us American.

Direct to Consumer Genetic Testing: What Makes Disruption “Sustainable”?

Genetic testing began as a niche service limited to a small number of specialized doctors that could test for a few rare diseases, but today consumers can test for over 1200 disease combinations ranging from breast cancer to cystic fibrosis. The falling costs and the lowering of technical barriers have transformed genomics and spurred a new industry: Direct to Consumer Genetic Testing (DCGT).  While doing some research on this industry, I found that the current practices of DCGT companies raise some troubling ethical concerns.This case study also inspired me to think critically about how “sustainability” is discussed in the start-up lexicon, and to propose a new definition of start-up sustainability going forward.

Perhaps the foremost critique of DCGT is lack of medical validity.  Contradictory results commonly arise from different tests, and the lack of regulatory framework means there is no quality control held over the science. Despite this, survey of 38 DCGT companies conducted by Leonhard Hennen in 2010 found that “only thirty-seven percent of the companies’ websites give specific information on the analytical validity of the genetic tests offered (accuracy of the test identifying the biomarker).” The concern is not so much that DCGT companies will misdiagnose a disease—government regulations prohibit diagnosis from anyone who is not a medical professional—but that a negative result  could cause consumers to think they are no longer at risk for a disease and delay a visit to a doctor.

Medical concerns also include clinical validity, or the relationship between the genetic marker and the patient’s clinical status.  In Dementia Entanglements in a Postgenomic Era, Margaret Lock uses Altzheheimer’s disease as a particularly illustrative example of the faulty risk assessments given on the basis of individual genetic testing. Although there was early excitement that the Alzheimer’s puzzle could soon be “solved” after the discovery of the APOE gene in 1993, scientists now unanimously agree that the presence of the e4 allele on the APOE gene “is neither necessary nor sufficient to cause the disease,” and that “at least 50 percent of e4 carriers never succumb to Alzheimer’s disease”. Scientists recognize that other genetic and environmental factor effect the onset of disease, but only 24% of companies in Hennen’s study give information on the clinical validity of genetic testing.

DCGT companies have also been criticized for failing to offer counseling services to those who use their services. In her study of deep vein thrombosis, Paula Saukko found that while some “well informed” patients felt counseling was unnecessary, a subgroup of diagnosed patients showed distress and confusion about thrombophilia.  This subgroup included those with a strong family history of DVT, a personal history of recurrent clotting, and who came from lower classes. The “less informed” patients felt that genetic counseling would have been helpful in dealing with their diagnosis, yet only ten of the 38 DCGT companies in Hennen’s study mention on their websites that they offer counseling service, and counseling in these instances is mostly done through written information via mail or web-log.  Of course, it is impossible to determine whether or not the customer properly understands counseling provided in those ways.

The start-up community loves to talk about “sustainability” in terms of financial metrics.  I have heard many times that if your Customer Lifetime Value (CLTV) is greater than your Customer Acquisition Cost (CAC), you have a “sustainable business.”  I think that going forward, we need to expand our definition of start-up sustainability. Sustainable businesses require not only sustainable financial metrics, but a sustainable market.  Likewise, a “sustainable market” generally refers to the size of the market (are there enough people out there who want genetic testing?). This is also an incomplete definition: a “sustainable market” comprises more than just market size. It includes the existential threat of regulation.  The American College of Medical Genetics has advised the public to avoid “home DNA tests” due to the possible misinterpretation of results and the lack of follow up counseling, and the FTC and the FDA jointly released a consumer alert targeting the lack scientific validity in some genetic tests. I am sure that the entrepreneurs who run DCGT companies are aware of the “best practices” espoused by their more traditional peers, and that they are aware that not following them has helped DCGT companies to raise massive rounds of financing.  But this growth is not sustainable because these companies have contributed in creating an unsustainable market.

DCGT companies argue that they give more consumers more access to the latest achievements of human genome research, and that by growing genetic testing supply they are helping to create a world of individualized, preventive medical care.  But DCGT companies will never reach their goal of improving health for consumers if they cannot first create a healthy industry.

“Solipsistic Startups” and the Downside of User-Centric Innovation

“Build What You Know” was the major lesson of the first session of the London Entrepreneur’s Challenge, a workshop programme and business plan competition open to a number of London schools that aims to show participants how to start a business. The lecturer, a former entrepreneur and venture capitalist who led the session, asked us to think of areas we consider ourselves experts in and talk about them with the people sitting around us.  The man sitting next to me told me he was an expert skier, and I informed him that I had sung opera since high school.  When we reconvened as a group, the lecturer told the 200 some-odd people in the audience that by virtue of our pedigree, we were some of the smartest people in the whole world.  This Challenge included Biochemistry PHDs and LSE MBAs, and if we home in on what we know, we can build amazing companies.

The lessons from that session mirrors what Eric Von Hippel termed the rise of “User Innovation”. As opposed to manufacturer-centric innovation, which originates from an individual or firm looking to sell a product, now the ones who serve to benefit from the product or service are the ones dictating the directing of innovation.  Collaborating users have done everything from inventing the World Wide Web to building the original mountain bike  because the existing solutions did not fit their needs.  Eric von Hippel and his colleagues discovered that “6.1% of consumers in the U.K. over the age of 18 had created or modified a product for their own use within the last three years,” and that rate is even higher among those with bachelor’s, master’s, or Ph.D. degrees in technical fields

While Eric Von Hippel neglects to talk about the demographics of the “user innovators” in his sample, it is clear that those who are able to turn their products into high-growth businesses have a certain profile. According to the University of New Hampshire’s Center for Venture Research,”minority entrepreneurs in the United States made up just 8.5% of the people pitching their businesses to angel investors in the first half of 2013″.  These entrepreneurs were also far less likely to receive funding, with only”15% of minority-owned firms successfully translating a pitch into dollar bills, compared to 22% of all businesses.” According to the latest CB Insights Venture Capital Human Capital Report, 81% of all funded founders are between the ages of 18 and 44 and 92% of these founders are male.  Less than 2% of venture-backed founders lack a college degree.

The result of this founder profile is an ecosystem of “solipsistic startups” built by entrepreneurs following the advice of “build what you know”.  C.Z. Nnamaeka published an article about this phenomenon titled “The Unexotic Underclass”, where she argues that entrepreneurs have not only failed to adequately address Big Problems like clean energy and climate change, but also big problems (little b little p) faced by people we ignore: single parents nearing poverty, military veterans waiting for medical benefits, and the unemployed who lack relevant skills for a modern economy.  While many entrepreneurs focus on how metropolitan twenty-something year olds choose restaurants and share status updates with friends, fewer focus in on problems faced by people they look nothing like or share little in common with.

What to do from here? 

For starters, we need to encourage entrepreneurship across a greater number of demographics.  If Eric Von Hippel and my entrepreneurship lecturer are correct, and “building what you know” and innovating on products you use generate the best outcomes, then the best kind of innovation in impoverished communities will require their members to lead the charge. One of the best models for encouraging innovation is the Israeli Yozma initiative, a government investment vehicle that combined attractive tax incentives with the promise to double any investment with funds from the government.  As a result of this program, “Israel’s annual venture-capital outlays rose nearly 60-fold, from $58 million to $3.3 billion program between 1991 and 2000,” and today Israel lists more companies on the NASDAQ than any other country outside of China and the United States despite having a national population just over 8 million. Perhaps governments could apply a similar public/private hybrid model to spur venture capital firms (whose returns as an industry are being beaten by the Nasdaq and S&P) to structure funds around underrepresented demographics (say, returning military veterans or minorities from low income areas).

In the meantime, I would reconsider the logic of “build what you know.”  Perhaps you love gaming and want to develop a new mobile game; unfortunately, nearly 60 percent of iOS app developers fail to break even with the apps that they create and market. About three out of four venture-backed startups fail: is that simply the natural rate of failure, or is the rate this high because similar entrepreneurs “building what they know” face a lot of competition (how many ways to book hotels do we need)?

In this blog post, Roberto Verganti argues that user-centered innovation has contributed in creating an unsustainable world. Verganti proposes a new solution, Design-Driven Innovation, which argues that innovation that originates from markets struggles to create new markets.  User-centric innovation tends to be iterative; design-driven innovation is transformative.  In order to truly create a breakthrough service, Verganti argues we should push beyond customers and users to find “interpreters,” visionaries who “deeply understand and shape the markets they work in.” Verganti gives the examples of Nintendo’s Wii and Apple’s iPod as innovation arising from this framework. Perhaps the best way to generate global utility is not to tell entrepreneurs to “build what you know,” but instead to ask them to build the future.