Govtech and My Experience with STiR

I am often asked about my work with Startup in Residence (STiR), a novel program run by the San Francisco Mayor’s Office of Civic Innovation to encourage and support startups tackling civic challenges. I’ll briefly describe “govtech" as a preface to explaining STiR and why it is a unique program for govtech entrepreneurs.

Govtech refers to technological innovation targeted at modernizing government and its services so it can better serve its citizenry. Governments spend $400bn annually on their technological infrastructure, which has largely been defined by legacy systems that promote inefficient processes and are incompatible with the modern interfaces to which citizens have become accustomed. Recent momentum in the space is driven by the realization that technology has empowered enterprises and consumers with a new suite of tools, but has not attended to the public sector.

My personal interest in govtech stems from its significant scope and potential for impact. Based on the figure above, a 1% efficiency gain from new technology results in $4bn of recurring savings. Surely productivity improvements are among the least controversial ways of increasing the efficacy of our public spend?

Perhaps, but acting on this opportunity has not been easy. Unfortunately, failed efforts to build govtech businesses have stigmatized the government as a slow-paced, budget-constrained, and often resource-intensive client. Consequently, entrepreneurs have elected to focus their efforts on other industries. Below is a list of some of the challenges that have hindered development in the space:

  • Procurement - it takes too long to sell into government. They may move too slowly for a sales force to be productive, and in some cases, the sale may be stalled further by a required RFP or tender process
  • Switching Costs - any single piece of software you might want to replace probably interfaces with several others (which also may be outdated); the switch may be impossible to do piecemeal and without large scale disruption
  • Requirements - even if a switch is possible, they may require a highly-customized solution, not the templated offering of a scalable SaaS company
  • Internal Resistance - people are afraid of change, are skeptical of the private sector, and do not want to lose their jobs
  • Budgets - many departments need to evolve their thinking on IT spend (SaaS vs. License + Maintenance)

However, these barriers to entry also highlight the opportunity in govtech. Governments are awaking to their need to adopt newer systems and re-brand their image as a client. They want to move quickly and understand that there are feasible approaches to implementing new technology. Entrepreneurs who are bold enough to test the integrity of their stance may be heavily rewarded. But where do they begin?

Introducing STiR. Effective solutions that solve real pain points will not emerge if entrepreneurs and governments do not interact, and we have observed that this gap is often too wide for either party to close independently. Structured as a 16-week program, STiR seeks to solve this by connecting departments within city governments to startups who want to build technology products addressing civic challenges. The success of the program hinges on the close collaboration between both parties and the subsequent symbiosis. The desired outcome for the departments is not only to co-develop a product they will want to consume, but also to learn about the iterative development methods of startups that might be effective internally. In return, startups receive direct interaction with a potential government client, regular mentorship, and access to the extended STiR network.

I am especially proud to support our team at STiR because we have converted this theory into tangible outcomes. Take our experience with Binti and San Francisco's Human Services Agency (HSA). HSA has many objectives, among which includes the support of foster care. They recognized that their parent identification and verification process, while valuable, was inefficient. The Binti team, equipped with skills desired by any tech startup, chose to work closely with the HSA staff to understand the intricacies of foster family placement. Binti and HSA subsequently co-developed a mobile-friendly, cloud-based solution that streamlined the process, making social workers 20-40% more productive and reducing approval times by 50%. The Binti team secured a contract with San Francisco, is expanding to counties across the state, and will have the potential to spread their results country-wide.

The potential for the Binti team to build a meaningful company is worth highlighting explicitly. Impact may not be a direct correlate of profit, but it does not exclude it either. Entrepreneurs who are passionate about improving government will find that there are significant opportunities to build large and sustainable businesses while also making a difference.

I look forward to contributing to our efforts at STiR in the years to come. We are just starting. Please get in touch if you are interested in learning more about the program.


Applications for the 2017 Startup in Residence program are open! Apply today to transform government through entrepreneurship:

Sneaky Price Increases

The holidays are prime-time for retailers to drive sales with timely product selection, marketing, and promotions. Leaving aside how companies get us to make a purchase in the first place, this season reminded me that pricing is in large part behavioral. What the consumer is able to bare really is as much a psychological question as it is an economic one. Companies are savvy and will always figure out the least intrusive way to increase the price on us, but how often do we notice?

While shopping for presents, I noticed that the discounts advertised were not as generous as presented. Many of the products had new list prices that were higher than normal such that the post-discount price was hardly a bargain. Yet I was drawn to the perceived discount and felt I was somehow cheating the market and getting a great deal. Anchoring is a powerful force.

Salience is another. Companies will frequently hold the list price of a product constant, and instead reduce what is included since volume is not as salient as price. This is effectively a price increase though. For example, the remote-controlled car that I bought last year is the same price even though it no longer includes batteries. This is common practice in the food industry where volume and contents are fungible. The average consumer does not notice that there is one fewer chicken tender or Doritos chip in a bag that cost the same as last time.

The last price increase I noticed this season related more to technology than to psychology, specifically that we accept small price increases if it is convenient for us to do so. We have a lot to be grateful for in the shift away from cash to credit and mobile payments, but there are some draw backs. Salience is at play since we are more likely to buy and accept higher prices when we do not physically see the money leaving our wallet. But beyond salience, the ubiquity of new payment methods allows price increases to be more gradual and convenient than ever before. The tennis club near my house has several vending machines. For as long as I can remember, water was $1.25 and soft drinks $1.50. It was that way for a while since raising the price by the only increment available, the quarter, would have caused a revolt. Now the vending machines are equipped with fancy credit card readers, and the prices are now $1.37 for water and $1.69 for soda. Perhaps I am okay with it since the credit card masks my actual cash outlay (salience), but I think it is more so the case that I was prepared for a price increase, but I just wanted it to be smaller and convenient (hunting for loose change is never fun). Smoother, gradual price increases are not new since credit cards have been around for a while, but since vending machines are largely behind the times, I experienced first-hand the coin-to-card phenomenon.

The list of our cognitive biases is long, so I'm sure there are many more techniques companies use to raise the price (and get us to buy things in the first place), so please share!

Case Studies and Blueprints

At a recent talk, we discussed misconceptions in business thinking. People offered compelling examples of how the dogma of the day did not apply to their experience. A common theme emerged: “I followed what X did, but it did not work for me.” I think case studies and other business insights are too often used as blueprints. They are not prescriptive and can be biased by the halo effect, a term coined by Phil Rozenweig.

Humans want answers. We want to understand not only what happened, but why it happened. But ‘why’ is often immeasurable, and we subconsciously use the nature of the outcome to assess the importance and value of the inputs. Starting with a known, quantifiable result (e.g. financial performance) biases our interpretation of how we achieved that outcome, especially when the inputs cannot be measured objectively. It's probably most effective to explain by example. Let's say you knew a coffee shop was doing well, and you were tasked with visiting the store to determine if and how the manager and her store layout contributed to the success. You are more likely to describe what you see as, “Cozy ambiance and dedicated manager who effectively supervises her baristas.” If instead you knew the coffee shop were struggling, you may describe the exact same coffee shop as, “Too packed, almost suffocating, and a micro-manager who slows down the entire operation.” It's the exact same manager and layout! With no metrics to quantify a manager’s efficacy or the effectiveness of a particular store layout, we let performance influence what should be a more objective assessment. If we can't measure it, we should not let "the halo effect" created by performance affect our conclusions.

The danger of the halo effect is not only that we interpret situations incorrectly, but also that these delusions creep into business thinking. We glamorize certain case studies and view their path to success as the path to success. We put their traits on a pedestal above all others. But this is not right. We do not have perfect information about what determines success, and over indexing on one example may dismiss many more important qualities that are not as salient. The recipe for a successful coffee shop does not mean it should be exactly like Starbucks. If we remember that outcomes are not perfectly correlated with inputs, we may recall that case studies and ’top 10 lists to success’ are biased; they can serve as convenient frameworks to understand examples of success, but they are not explanations of how to be successful.

Public vs. Private Research

Following the announcement of YC Research, several donations from billionaire philanthropists, and ongoing discoveries at Alphabet, our Dive discussion group explored the topic of public and private research. Here are some of the things we discussed:

Assorted Conversation Starters:

  • What type of research should the private sector fund? The public sector?
  • What is the distinction between private vs. public? Funding source? Whether results are made public or not? Who is conducting the research? Motivations of the research?
  • What type of research must be funded by the public sector? What types of things tend to be funded by the private sector?
  • Is one more effective than the other? How do you measure efficacy?
  • What types of goals should be associated with research?
  • Top talent is a scarce resource - where should they be spending their time?
  • Is failure more acceptable in one setting over the other?
  • Is it a good thing that more private companies are funding research?
  • What incentives exist for private sector research? Do patents make sense?
  • Do investors want their companies conducting research? How does the shift of Google to Alphabet relate to this?
  • Should tax incentives exist for companies that invest in research? That share their research?
  • Why does YC want to break into research?

Private Research (the points below were also up for debate)

  • More targeted
  • More efficient with spend
  • Greater immediate impact since market applications guide research areas
  • Researchers more focused; not distracted by teaching, tenure, grant money, etc. 
  • Positive externalities from creative research that affect overall company culture

Public Research

  • Step change innovation requires greater disregard for ROI and profitability, which public research can support
  • There is value independent of market potential to research; researching for knowledge’s sake
  • Public funding is required for disciplines and pursuits that are not easily monetized; there are many things the private sector would not fund
  • Private sector involvement can bias the output (e.g. Vioxx / Merck)
  • Public research often more collaborative with results made available, often eliminating redundancy of work; not true of research in the private sector where findings are held as proprietary insights
  • Can tolerate more risk
  • Better access to information since public researchers are unbiased 3rd parties

Google's Transition to Alphabet

  • Greater need for transparency 
  • Have any moonshots actually worked? Efficient use of cash?
  • Google is large enough to form Alphabet and has convinced investors that the moonshots are worth it; can other companies position themselves similarly?


  • Supposed to encourage innovation, but the system is outdated in many ways
  • Patent trolls
  • Patents require disclosure of invention - but descriptions often obfuscate essential facts, limiting ability for incremental innovation
  • Risk of being sued for IP infringement (unnecessary money spent on defensive and offensive tactics)
  • The requirements for what constitutes 'non-obvious' should be strengthened
  • Patent terms are too long for some industries 
  • 40-60% of patents are "un-used" - should there be a 'use it or lose it' approach?

Purposeful Business

Warby Parker, Toms, and several other well-known consumer facing companies have emerged with purpose at the core of their missions and cultures. Meanwhile, the common objective of business is to maximize profit and shareholder interest. Does an intersection between the two exist? This was one of many questions raised at a recent group lunch on purposeful business. It was a fascinating conversation with much more left to discuss, so I wanted to share a few of the interesting areas that we explored:

One person made the point, which I agree with, that industry structure and profitability define a company’s ability to “give back.” In a competitive industry with thin margins, companies cannot afford to give anything away. Conversely, companies in industries with monopolistic characteristics (e.g. eyewear) and/or high margins (e.g. branded apparel) can redirect dollars to a purposeful cause.

But what does "giving back" even mean? Companies in highly competitive industries can still pursue purposeful missions if consumers preference such behavior. Purpose, therefore, would not come at the expense of a company’s bottom line. If this is true, then ‘purpose’ is a sound business decision and 'giving back' is a smarter, more profitable strategy. Purpose provides differentiation in an otherwise commoditized industry.

To what extent do the monetization of and strategy behind ‘purpose’ blur the distinction between profit and purpose? It is hard to think of companies that have not benefited from adopting a social mission. Though difficult to measure, the purpose initiatives of Warby Parker and Toms surely have helped them grow. Relatedly, how should we understand the difference between a B-Corp and a C-Corp? Is it that they have different objectives or that the source of their differentiation and strategy are of a different nature?

How does the monetization of purpose affect our understanding of what is authentic and what is not? Does the distinction between authentic or inauthentic purpose matter? Warby Parker and Toms are paragons of the B-Corporation. They are viewed as authentic brands, in part, because there are so few examples of companies that do this. I think their authenticity emerges because profit is not the sole driver of their purpose. They certainly have monetized purpose, but it seems to go beyond pure profit. I have defined authenticity as not the opposition of making money, but rather as the motives underlying purpose. Disaggregating motives behind corporate initiatives is hardly a science, which is why branding and communications will have heightened importance in the profit-purpose world. For better or worse, customers have an innate distrust of big companies, even if their startup counterparts have the same motives.

If consumer behavior shifts to favor socially-conscious companies, companies will accordingly adopt social missions at scale. How does ubiquity change our understanding of authenticity?  

How do we measure the impact of a social purpose driven strategy on a company’s bottom line? By how much did it increase sales? How did it improve employee retention?

Should 'purpose' driven companies receive preferential tax treatment? How do we define purpose? Is it relative? That which may have been purposeful a decade ago may be mundane today. 

What role does regulation play in nudging consumer behavior?

Price and Value in Venture Investing

This post was motivated by a few conversations I had with public market investors in which we explored whether value investing theory could be applied to venture / growth capital markets (forgive us for nerding out and ignoring the more practical indicators of what makes a good seed/growth deal). I have hand-picked a few of the more interesting questions we discussed.


As background, the efficient market hypothesis says that the price of every security (e.g. a stock on the NYSE) represents all available information (is 'efficient') and is 'fair' since all returns are related solely to risk. There is no free lunch since the market knows everything. However, for a variety of reasons (e.g. human psychology, poor analysis, etc.) the price of a security can deviate from its intrinsic value, thus creating opportunity. A value investor believes that he/she can only generate a return in excess of what is considered ‘fair’ if he/she has a view on intrinsic value that is differentiated from consensus (price). If less than the price, the company is overvalued (according to the investor), and if greater, than the company is trading at a bargain.


Does deviation from price vs. value thinking contribute to the high price environment / possible tech bubble?

Part of our conversation focused on the sources of high valuations today. One person argued that venture investors extend valuations to companies far in excess of their value, thereby contributing to tech bubbles. Perhaps it is semantics, but, absent strategic rationale, I don't think investors are participating in deals where price > their opinion of value (e.g. it is not worth $500MM, but let's invest anyway). I suppose all investors are technically 'value investors' if they formulate some personal view of what the company is worth and act rationally when comparing it to the price. Instead, I think prices are high partly due to 1) investors underestimating risk / believing a lot more and 2) investors underwriting lower returns. On the former, price will rise when investors underestimate risk, which is bound to happen in bullish times when risky investments have panned out. It is not sufficient to look at the 'unicorn' outcome of expensive deals and conclude that the decision to invest was correct. Either those investors were exceptionally bright or luck was in their favor. Impossible to say which is which without looking at long term track records. On the latter, investors may be willing to accept lower returns for a variety of reasons:

  • Interest rates are low, so LPs may be willing to accept lower returns for all other asset classes
  • There is greater competition among capital sources (everyone wants to invest in startups), and not all are smart. Momentum investing is common in bullish times (the company is growing, therefore it must continue to grow); these investors probably do not have any view of value and are investing just because a deal is 'hot' with the hopes that it continues to be hot. All of this money pushes up the price, and as a founder, why not take advantage of it?
  • Access to hot deals is prized as the ultimate trophy in many ways by VCs and LPs alike. Whereas many investors in other asset classes are okay missing out on the home-runs just as long as they have no losers (i.e. downside protection is a priority), venture is in many ways the opposite. The penalty for a 0 is often less severe than the penalty for not being in a hot deal. LPs like to see that their funds were in Uber and Twitter. Additionally, VCs gain a lot of street-cred based on the logos they boast are in their portfolio. Access is king, and to the extent that it puts investors in the 'we need to get in at any cost' mindset, it can add to the pricey environment. Not a problem if the company lands up being a unicorn, but not all companies will.


Taking a step back, is price even important in venture and growth investing?

In theory, all else equal (e.g. cost of capital, synergies, value-add, etc.), the prices that two VC firms set should be the same unless their views on opportunity size or riskiness of the deal diverge. With such a wide range of possible outcomes (for example, anywhere from $0 to $100BN), it seems impossible to set a precise value for a startup. With the variance in outcome so large, the implied intrinsic value is likely to be a large range as well (the inability to actually determine value is why it is hard to really describe VCs as 'value investors', though in theory they are). With that, how do we explain this recent (true) example? A VC firm was willing to value a company at $45MM, but saw the $60MM offer from another VC firm as crazy and walked away from the deal. Do they really have so much insight that they can value a startup so precisely? Assuming they are targeting a 10x, do they have so much conviction that the deal will eventually be worth $450MM but not $600MM?

From my conversations, I think the decision to walk away ‘based on price’ has less to do with price vs. value, and more to do with desired exposure and ownership targets. A VC firm wants to have $10MM of exposure to a startup, and wants to have 20% ownership, therefore implying a post-money valuation of $50MM. If another firm comes along offering $20MM for 25% (implied post-money of $80MM), the first firm might walk away. Was it really because the first firm could not imagine a scenario where the company could be worth $800MM? Unlikely. But if it wanted to invest with the same ownership/governance goals, it would now require more exposure, which may not be something the VC firm has the appetite for. My simplistic example seems odd since the valuation is not dependent on any business fundamentals, but I'm assuming that is incorporated into exposure / ownership considerations.

To the VC folk, is this consistent with your experiences? What other reasons might you walk away from a deal 'based on price'? 


It is only worth doing diligence on a company if you have a differentiated viewpoint.

Public market investors typically will only spend time analyzing a company if they believe they can develop a differentiated perspective. If they rely on consensus to inform their view, there will be no difference between their opinion and market price; therefore, no opportunity exists.

I don't think this statement carries to the venture / growth market. The requirement of a ‘differentiated view’ is predicated on the fact that the price is set by the market, is more or less efficient, and represents consensus view. In the venture/growth context, the price is often driven by non-market forces. For example, entrepreneurs regularly choose investors for non-economic reasons (better fit, the right connections, other terms, etc.). Access is another reason; 'extra' demand in the public markets will drive the price up, whereas excess demand in the private markets often means you are not getting into the round. For these reasons (and many more), the price may not represent supply and demand. Therefore, as a potential co-investor in a deal where the valuation is already set, I am of the mindset that unless you cannot understand what the company does, you should be open to doing diligence on the deal. The lack of a true market price, in other words, negates the aspect of value investing theory that requires a differentiated view. Note this is ignoring the practical limitations investors might have on ownership, governance goals, exposure, etc.

The Future (Gore)

The Future by Al Gore

Key Points & My Random Thoughts

+ The developed world has been losing jobs to two trends: outsourcing (to countries with lower input costs) and robosourcing (to technology/robots). Emerging countries may benefit from the former trend, but many in the developed world are suffering since jobs are not being replaced. This is endangering the middle class. Gore argues that wealth should be redistributed to compensate for unemployment related to these trends. Otherwise, income inequality will continue to grow.

In theory, the world as a single system should be better off if everyone focuses on their core competencies. In theory, as robots and other technologies replace jobs, human capacity is available for more value-added activities. In the long run, the growth of these trends will likely be a net positive. Even though the machines of the industrial revolution put people out of work, few would contend that this was bad in the long run. After all, we all now enjoy cheaper products, have a wider breadth of services, and have freed up time to innovate elsewhere in society. The long-run picture may be fantastic, but we cannot  ignore the short-term reality. In the short-term, people lose their jobs and their livelihoods to these trends because of frictions in the market. Unfortunately, neither people nor society are not able to adapt as quickly as the trends develop. The pace of technology and robosourcing, for example, will evolve faster than anyone's ability to acquire new skills, generating a skills gap resulting in unemployment. Rather than halt innovation and oppose technology, we need to think about how to accelerate skill development, facilitate job transitions, and provide strategic and financial support in the short-term. Assuming there is a net benefit overall to these trends, society will adapt to accommodate them in the long run. (My thoughts assume just a single system and ignores the fact that there is an uneven distribution of resources and opportunity across socioeconomic classes, across countries, etc., which could mean greater income inequality… but that is different than consumption inequality).

+ Global power rests with major corporations, not the United States. American politics panders to the corporations that support it and candidates who care about attaining and maintaining their positions must listen to their constituents.

Not much to say here since I agree with Gore. The election process, which is heavily dependent on campaign financing, is over-influenced by major corporations and those with money. Tough to blame anyone since political candidates really do need the cash to promote their name and generate national recognition. But how can we shift power away from money? It seems that money carries weight because we are, in large part, passive citizens who learn about things when they are paid to be shown to us in the form of advertisements and lawn signs. If we want to change the paradigm, we need to be active. We need to do research, discover the candidate that refuses to make deals with major corporations in exchange for money, and help elect her/him. Does the fact that we have not taken this active approach reveal a subconscious state of contentment? Perhaps we do not understand the consequences of being passive citizens? Perhaps the tools for discovery and promotion are insufficient? I think it is less the latter and more the former two.

+ The benefits of the Internet are several. It has improved our lives in inestimable ways and promotes democracy. However, it can also be used for nefarious purposes.

No doubt the Internet is improving our lives, but is it promoting democracy? It has been a powerful tool in exposing injustice, especially in the last several years. It gives volume to voice, but does it give power? Whether it helps promote democracy depends on whether or not this voice actually influences political decision making. If it is not changing the outcome of our political processes, well then we have room to improve. Separately, while the Internet can be used to promote democracy, we cannot forget that it can also be leveraged as a tool for autocratic regimes to censor and surveil its citizens.

+ There are significant opportunities and threats in biotechnology, specifically agriculture and medicine. Genetically modified food, for example, can help provide a sufficient supply of nutrients, but may also contain health risks. Similarly in medicine, while there are benefits to using biotechnology to intervene in health disorders, the same techniques can be applied in negative or controversial ways (e.g. gene selection and “playing God”).

I discussed the dualistic nature of technological developments (in that they offer both benefits as well as drawbacks) in an earlier post ‘Soap or Dirt?’ With many technologies, we will not know beforehand whether the net benefit will be positive or not, which is a concern but does not mean we should halt any and all efforts to make progress in a particular space. I tend to skew towards exploring what technology can offer, and before promoting and scaling the solution, testing it with a small yet significant sample size. It’s tough to approach every technology in the same way, but at the end of the day, this requires openness on all sides.

+ Population growth accelerates the consumption of already scarce resources. By 2100, could we reach 10 billion people? Urbanization puts additional strain on resources as well as on governments. Technology alone cannot help us. We need policies.

I do not have a strong view on population growth but I agree that policies can be a good hedge against technology risk. It’s tough to imagine betting our planet on someone inventing a cure some day. In instances of such uncertainty, policies that provide incentives to curb potentially harmful behavior may be helpful. With that said, I think new design / tech-enabled companies can help shape our behavior just as, if not more, effectively than top-down mandates from the government. Mitigate technology innovation risk with behavior innovation.

+ Over-emphasis on consumption based metrics (e.g. GDP) leads to short-sightedness. To adopt a longer term view, we should include environmental and other important variables into our measure of health and productivity.

I thought this was a good point. GDP is a production or consumption metric that, to my knowledge, does not consider the consequences of production techniques or consumption trends. It is very possible that unsustainable practices boost GDP in the short-run at the expense of long-term health. It could be interesting to weight GDP to account for future wealth and health. Perhaps the present value of future GDP, for example. If you over-use resources and engage in unsustainable practices, while year 1 GDP may be higher, the PV of all future years of GDP may be lower. How you forecast future GDPs in a logical, consistent, and agreeable manner is a much more difficult problem.

+ Climate change continues to danger our world and the onus really falls on governments to stop corporations from engaging in deleterious activities. We need an 80-90% reduction in CO2 emissions.

Companies need to think about and internalize the negative externalities of their actions. Governments can mandate this, but we as consumers are powerful changemakers as well. If powered with the right information, we can start selecting greener products, even if they are sold at a premium to non-green alternatives in the short term. I am not sure what the data is to support whether consumers actually believe this though. I imagine in tough economic times consumers are more price sensitive than socially conscious.

Emotional Intelligence (Goleman)

As a quick preface, I am creating a category called 'Books' and plan to share the key points from and my thoughts/reactions to some of the books I read. Hoping these help start some discussions.

Emotional Intelligence by Daniel Goleman

Key Points
+ Emotions have several practical benefits: 1) they provide context to facts and experiences (e.g. action X typically results in sadness), 2) they enhance our ability to interpret the feelings and actions of others, and 3) they mold our instinct and intuition.
+ Emotions have drawbacks too: 1) they can make us act impulsively (neurologically, information bypasses the thought region of our brain (neocortex) and goes directly to the emotional center), and 2) they cloud our judgment when, for example, we are in a heightened emotional state or if we have been ‘emotionally scarred’ from an event (“residual obsolete emotions” is the more technical term used in the book; an example might be if you are irrationally afraid of small puppies because a big dog happened to bite you when you were a child).
+ Control and understanding of emotions and how they make people look, feel, and act constitutes emotional intelligence (“EQ”). Those with high EQ are great at reading and navigating social situations. They have a great balance between the emotional and the rational, the feeling and the thinking. Emotional self-regulation describes the process of controlling our emotional brains with our thinking brains.
+ EQ is important, yet it is under-emphasized in education. You can improve your EQ by being observant to both your personal as well as others’ emotions.

My Wonders
Emotional intelligence is a fascinating concept that Goleman helped pioneer. I think one of the more interesting points is the mention of emotional intelligence in school curricula. Wired recently ran an article on the growth of home schooling, especially among the tech community. It suggest that the growth is driven, in large part, by the desire to avoid pushing students through programs where standardization is required to excel and creativity inadvertently is sidelined. We can debate the merits of that separately, but to merge the concepts here, does home schooling hurt the development of emotional intelligence? Should traditional schools emphasize emotional intelligence in their curricula? What solutions exist to augment curricula with such education? Edtech trends are distributing academic content and tools for analyzing student performance at extraordinary rates. But do any emotional intelligence tools exist? What exercises actually help build EQ? Perhaps this line of questioning is missing the point. Maybe it can’t be a tech solution. Maybe EQ is developed as you enjoy life, hang out with friends, watch movies, and are observant as you live. Perhaps the best product to enhance EQ is no product at all. Or a pen and personal journal.

Is ROI in Education Too Simplistic?

The concept of ROI in education is not new. Majors have been ranked on this metric for years. Engineering and economics rise above gender studies and english literature in most analyses, while Ancient Languages and Medieval History are viewed as a courses of study entirely too anachronistic to even consider in the reports. Putting aside vocational schools that market to students the promise of professional skills and income generation upon graduation, I think ROI based on income is too simplifying a metric in education to be prescriptive. Just because income is measurable does not mean that it is the only useful metric, and presenting it as such is potentially negative. ROI, a measure of income generation vs. costs of education, is incomplete because 1) it understates value creation and 2) excludes non-financial variables that matter to us.

On the first point, income is not synonymous with value creation. Your income is likely to be higher at a corporate law firm than working in a health clinic in rural Africa, but is it multiples more valuable? For a variety of reasons, certain careers are not lucrative (e.g. industry structure and relative power, profitability, positive externalities cannot be internalized, etc.), but we cannot ignore that there exists value that is not captured in income (or gets captured in somebody else’s income). ROl as published ignores the fact that writers may influence the creative thinking of engineers even though they capture only a fraction of that value in the form of book sales. It will undervalue historical linguists whose work may actually inform modern day natural language processing and encryption. A response to this could state that value generated for society does not pay the bills and therefore is not a consideration that people practically build into their personal calculus. Value generated for one’s self is exactly equal to income. That is not necessarily true, and this takes us to point two.

There are non-income related sources of ‘return’ that are very much part of personal decision-making on 'school vs. no school’ or choice of major. Exposure to diverse perspectives, the experience itself of living in a community of amazing people for 4 years, meaningful relationships, personal fulfillment and happiness from one’s career, acceptance in one’s community are all examples of important characteristics not always captured in income. If income were directly proportional to personal fulfillment, would the distribution of wealth across professions be the exact same as the status quo? Likely, no.

Now I am not arguing that ROI as we know it is a useless metric. The cost of education is rising and student debt is a very real concern. Income is important to consider when choosing a career because it can affect the life you want to live. What I am saying is that it is not the only thing to consider, and should not be presented as the be-all and end-all metric. Over-emphasizing ROI when considering the value of education or of particular majors could result in some sort of unexpected distortion. Exaggerating for the purposes of making the point, if everyone chose their majors based on the ROI income data, we will over index the population to skill sets and careers that are able to capture the most value as income in a single point in time (the present) at the expense of careers that actually generate the most value (independent of income captured). We may also lose the diversity of thought and skills sets that are required to innovate and advance over long periods of time.

I will conclude my ramble by saying that publishing the typical incomes associated with each major is probably a net positive since it paints a realistic picture for students as they make important decisions. But to ask questions like, “Is education worth it?” or “Is studying English worth it?” while exclusively citing this metric misrepresents the full story and discounts many important sources of value and fulfillment.

Machine Intelligence: Discussion Points

Our discussion on machine intelligence spanned short term opportunities and long term disaster scenarios. Some of the points we discussed are included below.

Intelligence is not only defined by the speed with which answers are determined, but also the quality, scope, and nature of the cognitive skills. 

The application of machine intelligence to every industry is a hot theme across organizations of all types. The sudden adoption of it may not be realistic however due to limits on technological capability and human acceptance. Technology will improve over time, but will psychological barriers (e.g. how could a computer drive my car?)? Are younger generations who have grown up around technology more welcoming? Are there certain fields (e.g. political decision making) that we would never want a machine doing?

Companies have tried testing customer appetite for machines in low fidelity ways such as having a human do the processing behind the backdrop of an 'intelligent machine.' Beyond being an efficient way to test a concept before deploying resources to build a product, it also helps collect data that can be used to train the machine and develop data sets.

In charting the adoption of machines, we have moved from human-only to more machine-assisted / human-assisted approaches. For example, a plane is in large part auto-piloted with humans there to assist where necessary. Will this shift to purely machines eventually? Will we accept this? How do we accept this? Will someone have to prove that the error rate for machine-only approaches is lower than machine+human? How does one actually conduct that test?

Another axis to evaluate adoption is based on complexity and harm. High volume tasks that are low risks seem ideal to tackle first. High volume allows for sufficient machine training, and low risk implies that machine errors have minor consequences. Scheduling meetings and booking travel are examples of applications that meet these criteria. 

Existing approaches seem to be very narrowly defined, with programs applied to small micro tasks (e.g. travel booking, calendar planning, etc.). These are difficult to build comprehensively, and so there is likely going to be a growth in companies offering one-task solutions. Eventually it will be difficult for a consumer to manage these separately, and “dispatcher” layers (e.g. command centers like Siri) will be built on top.

Talent will be distributed across small startups, large tech companies, in academia, etc. Big companies do have advantages here, including ability to poach and pay top talent, willingness to fund research, access to necessary data sets, competition and a business need to develop machine intelligent solutions, etc. Big companies are likely to be hot spots for this activity, and they are trigger happy on acquiring new companies in the space ('acquihire'). As an investor in one of these startups, you may determine whether to accept or reject the offer based on the standalone prospects of the startup (e.g. can and are they making money?).

From where will general intelligence come? Will it be from a team that is working on cracking general intelligence? Could we accidentally come across the discovery? Imagine someone were building an intelligent machine for an application in agriculture, for example. Could his machine unexpectedly be an expression of general intelligence? Perhaps the elements required to unlock this are not all that far off from where we are, in other words.

Intelligence is not just about the 'thinking' or 'brain' aspects. It requires sensing and learning. It requires the 'body' as well.

Machines and machine intelligence will replace jobs? Will those jobs be replaced? If so, with what? If not, what are the consequences? What does automation and machine intelligence imply for the nature of work? Free time? Social life? Family dynamics?

Long(er) term outlook: Are the threats of superintelligence fact or fiction? There is no consensus about the end state but many concerns are valid and should be openly discussed. Machine intelligence is currently limited to niche applications, but as it develops into general expressions of intelligence, the rate of its development will skyrocket and almost immediately become superintelligent. While machine intelligence promises the hope of cracking the code on difficult problems we need solving, computer programs with great intelligence may dangerously consume resources when solving problems. Will machine intelligence necessarily evolve exponentially? Are we ignoring the fact that the difficulty in 'unlocking' the next step of innovation also rises over time.

Four Tendencies

Gretchen Rubin’s new self-help book Better than Before: Mastering the Habits of Our Everyday Lives describes habit formation in the context of four tendencies or personality types. Rubin argues that awareness of one's own type can help with the efficacy of the habit forming techniques presented in the book. While her classification can catalyze honest introspection, it can also help us understand others in a meaningful way. For that reason, I wanted to share a quick summary and see if others had come across other categorizations that they found interesting, accurate, or at least thought provoking.

Rubin’s four tendencies revolve around sources of expectations and our responses to them. Expectations are of two types: external and internal. external ones are imposed by others (e.g. your boss gives you a deadline, your mom assigns you chores, etc.) whereas internal ones are self-imposed (e.g. a New Year’s Resolution, a personal goal to diet, etc.). 

1) Upholders (External and Internal) respond well to all sources of expectation. They please authoritative figures, satisfy deadlines, and are self-motivated. Unfortunately, they may also blindly follow rules and may inadequately question tasks assigned to them. They can also struggle in environments where expectations are unclear.

2) Questioners (Internal Only) are driven solely by internal motivators. They are self-driven and reflective. They do not ignore external tasks, but only engage if they deem them purposeful. They analyze external assignments and determine whether they make sense rationally to complete before proceeding with the work. Rubin describes questioners as people who question everything and use careful assessment to convert all expectations to internal ones. They can often ask too many questions (annoying) and analyze tasks too heavily, which slows processes down.

3) Obligers (External Only) struggle with self-imposed expectations but are eager and quick to please others. They would sooner help a colleague with a task than work on an internal goal. They have every assignment completed at work, but routinely ignore personal goals.

4) Rebels (Neither) are rare since they approach goals in unorthodox ways. They resist external expectations and, while they can have personal goals, will not set internal expectations in traditional ways (e.g. New Year's Resolutions).

While most things psychological and behavioral cannot (and should not) be put into boxes, attempts at identifying patterns and offering simple taxonomies can enhance our perception, interactions, and sensitivity for others. For example, it is probably ineffective to let obligers create their own goals and have flexible, self-directed schedules. Or it is probably difficult selling a product to a questioner without explaining the the value proposition and core differentiators in full detail.

Soap or Dirt?

Books like Superintelligence by Nick Bostrom have made graphic the potential calamities of developing machine intelligence. A cautious-to-pessimisitic view on the impact of technological progress is not new, and even the strongest of technophiles among us have to agree that there exists a chance for unintended consequences and deviations from intended use with any technology. Unfortunately, there is always dirt in the soap.

This theme, that all good comes with some bad, is a very natural phenomenon. It is ubiquitous, in our careers, our relationships, and our biology. The Economist recently ran a piece on how the very source of our mental capacity, in certain cases, can cause Huntingdon’s disease. I admit, I sometimes fall into the mode of thinking that all things natural are somehow error-free, designed with omniscient purpose or adapted to perfect form over centuries of survival (of the fittest). That is not true, good enough to last is not synonymous with perfect.

If the research the Economist cites is correct, I’d be willing to bet that most people would not trade the intellectual power of our entire species for a reduction in the risk of Huntingdon’s disease (though just typing that makes me feel harsh and unempathetic - I assure you that is not the case). What does this say about the arguments of the technophobes? Is it unnatural and abnormal to fight against artificial intelligence? Do they ignore that technological advancement simply mimics how biological enablers of our success have drawbacks? 

While it is interesting to note the parallel between technology and biology, the two are very different and the questions above are misleading. Technology advances at a rate much faster than does biology. Genetic deviations introduce incremental change that gains broader prevalence gradually if and only if it exhibits more favorable survival characteristics. It’s hard to see how any single change would wipe out an entire species, whereas that is not true of technological development. Technology exhibits exponential growth that can outpace our ability to react.

Our inability to quantify the ‘good’ and the ‘bad’ is why this debate exists in the first place. So don’t dismiss books like Superintelligence entirely. Discussions about the future state of our world and the role of technology are fascinating, diverse, and worth having. While we all may agree that no soap is pure, we need to know when we should just start calling it dirt.

The Next Billion: Discussion Points

I was recently part of a fascinating discussion about what has been termed 'the next billion' describing the next wave of people to come online (predominantly from emerging markets) and the opportunities/challenges this presents. Sharing the main points below: 

There are several issues we discussed including connectivity, access, devices, usage of devices and data, monetization strategies for apps, ROIs for different players in the space, etc.

+ Connectivity: Balloons, mesh network of buses (avoids real estate issues), satellites, etc. What return can be expected from these investments and will they be successful? Internet is a utility, and charging for infrastructure may be difficult. Therefore, is the set of solution providers limited to large companies (e.g. Google, Facebook) with the capital resources to fund connectivity at a loss in exchange for some hope/ability to monetize subsequently in the form of apps?

+ Devices: Costs matter in this market, hence the race to deliver lowest cost devices. Will Apple be sidelined? Can they avoid brand dilution while foraying into these markets? Is open source more likely to win than closed systems?

+ Usage of devices and data: Usage is very different in the developing world for a couple of reasons. First, there is less income to spend on data and in-app purchases, affecting engagement and product design. Should all apps be lite versions of their developed world counterparts? Can we utilize cheaper ways of downloading or sharing data?

Second, because of the lack of financial infrastructure (e.g. credit cards), the carriers hold the keys. Identification and payment occurs at the carrier (e.g. when phone is issued, when minutes are topped up), thus app stores are carrier controlled. Implications for Google and Apple? Implications for apps’ ability to monetize (opportunity costs are not just financial, but also the hassles associated with running out of minutes). Financial infrastructure is very much an important driver here. mPesa is an example of carrier (Vodafone) power in Kenya.

+ Monetization: A few ways to pay for apps: money, usage (e.g. data), attention (e.g. ads). Ability to charge for apps / in-app upgrades / use significant data is limited due to limited disposable income. Attention is premised on subsequent ability to monetize. Advertisers effectively subsidize data costs on behalf of the consumer (the same thing happens here) in exchange for leads and purchases. But the returns have to be realized at some point. When? Will users with limited disposable income respond to heavy advertising in the same way?

+ Correlation between GDP / GDP forecasts and the value of a user. “A user is a user” is not true. Developed world users are very different than developing world users. Even within a country this is not true depending on how wealth accumulates and is distributed. Per user metrics (related to valuation and otherwise) should adjust for this.

+ At first, many companies are likely to focus on logistics/infrastructure (e.g. the Uber / Instacart type companies) in emerging markets than heavy advertising models.

A Knights and Knaves Riddle (Part 2)

This is the second part the Knights and Knaves riddle that I heard recently. Check out Part 1 if you have not already. Good and necessary warm up.


You are traveling and come to a fork in the road. One path takes you to your destination and the other to your death. Unfortunately, you do not know which is which. Luckily, three people are there to guide you, a knight, a knave, and some indistinguishable person whose responses are entirely random. You do not know who is who, but you are allowed to ask two yes-no questions to any of the three men that you please. What two questions do you ask to reveal the correct path?



[To be posted soon... many of you told me you wish I had not included the solution immediately. Why you were unable to stop yourself from reading more is a separate riddle.]

Rationale for Tech Acquisitions

Acquisitions play an important role in the tech ecosystem (and unfortunately, can also contribute to the bubble). How do acquiring companies think about what and what not to buy? What is their rationale? Using a few consumer acquisitions as a starting point (e.g. WhatsApp, Instagram, Vine, etc.), I am compiling a working pro-con list from the acquiring company’s perspective. Please add!


Acquiring company’s ability to accelerate target’s growth / engagement and vice versa

Defensive tactic to avoid disruption or eliminate a threat

Better monetization ability and/or tactics

Improve user experience by merging products / workflows

Market power, greater network effects with larger user base

Access to a more valuable user base

Diversification of offerings

Block a competitor from acquiring the target

Inability to replicate the target’s product internally

Enhanced data collection

New technology and competencies (e.g. Oculus for Facebook)

New Talent (acquihire)

Improve target’s processes and/or eliminate cost redundancies (cost synergies likely limited for tech startups though)

Target company mismanaged












Difficulty pricing the target (e.g. uncertain pro forma growth, monetization, profitability, etc.); high valuations and uncertain ROI

Chance that the incentives of target’s employees are changed post-exit

Vision misalignment and/or improper culture fit (keeping the company independent may help mitigate this risk, e.g. Instagram)

Privacy concerns as companies and their access to personal data scale

Cheaper to build and launch the same product internally

Unrelated diversification; distraction from core goals

A Knights and Knaves Riddle (Part 1)

A riddle recently went around work that I later learned was a 'Knights and Knaves' logic puzzle since it involves knights, who exclusively tell the truth, and knaves, who are chronic liars. Below is the first (and easier) one.


You are traveling and come to a fork in the road. One path takes you to your destination and the other to your death. Unfortunately, you do not know which is which. Luckily, two people are there to guide you, a knight and a knave. You do not know who is who, but you are allowed to approach one of them and ask a single yes-no question. What question do you ask to reveal the correct path?

Solution (yes, spoiler alert):

Approach one of the men (it does not matter which). Point toward one of the paths and ask “Would the other man say that this path that I am pointing to is the correct one?" This question works because both men will give the same response. How?

When asking the question, let’s assume you are pointing to the correct path.

The Knight's Logic: You are pointing toward the correct path. You have asked if the other man, the Knave, would tell you that this is the correct path. Well, he only lies, so NO, he will lie and tell you it is the incorrect one.

The Knave's Logic: You are pointing toward the correct path. You have asked if the other man, the Knight, would tell you that this is the correct path. In fact, he would, since he tells the truth, so I must lie and tell you NO.

This shows that if you are pointing to the correct path, both will respond NO.

By the same logic, if you are pointing toward the incorrect path, both will respond YES.

In this way, a single yes-no question will give you sufficient information to reach your destination.

This was prep for the second riddle, which is harder!

Is binge engagement bad for retention?

The human behavior and psychology underlying product design can and have been described through nearly every lens, but I recently heard it discussed in the context of food, which I found refreshing and new. Frank Mars, the founder of Mars Candy, believed that a perfect candy bar left the consumer 'wanting one more bite.' The overly decadent chocolate cake that, while delicious for the first few bites, quickly becomes the source of regret and discomfort is a product whose design failed to incorporate optimal portion size. To Mars, controlling consumption was integral to perfect product design.

Does this principle apply elsewhere? To tech products? Are there certain apps whose value and appeal could be enhanced if their use were limited (e.g. 5 minutes a day, 10 actions a day, etc.)? Intuitively, it would seem odd for a company to limit the use of its products based on this principle alone unless it is true that, like candy, overindulgence at time period 1 results in lower retention or use of the product in future time periods. I think this principle probably applies best to products that are enjoyed and/or have some addictive properties about them (e.g. many consumer apps as opposed to enterprise ones).

So would this principle work for a social media site, for example? Most of these products create a ‘hook’ based on the dynamic nature of the content, something new to look at with every visit (e.g. new pictures, new posts, new connections). They rely on intense engagement for content creation, which in turn enhances future engagement, generates value for one’s network, and supports monetization. It seems odd then to limit consumption as intense engagement results in a better product experience. So the answer is probably no.

What about for Snapchat? We note that the ephemerality that is at Snapchat’s core is a very different type of limitation on consumption, but would a Snapchat that only allowed you to take 3 snaps a day be interesting? What if it only allowed you to open a subset of snaps sent to you? Hard to argue that these changes would enhance the product experience, but what are examples of products for which this principle applies? Are there certain properties about a product that lend itself to enhanced retention and enjoyment due to controlled consumption?

Perhaps the products, like candy bars, have to be the same across time periods. Regardless of whether you eat a candy bar today or tomorrow, it will taste the same. The sites discussed above offer different content every day so it makes sense that they don’t fit this principle, at least at first glance (it would be an interesting experiment to see something with consumption control). What about gaming? Many games offer the same experience whether played today or tomorrow. I certainly have binge played a game to the point of exhaustion. Had my engagement been limited in the beginning, would I have been a more frequent and valuable user to the gaming company in the long run? I am not sure what the LTV of binge users who drop off (‘hares’) vs. slow and steady users (’tortoises’) is, but I wonder if gaming companies have evaluated this. I admit it would be an odd move to restrict users from consuming your product, but is that intuition grounded in myopia rather than long-term product analysis?

Do you think this principle could apply to certain products? Have you seen it applied at all? Successfully?

Some food for thought.

Disruption Theory in the Food Industry

In a recent conversation with a food executive, we discussed an aspect of disruption theory that I have only really heard about in the context of technology. In the consumer food industry, large players make products for the masses, as economic intuition would suggest. If it takes the same input costs to produce a product that millions will enjoy as it does one that thousands will enjoy, make the one with broader appeal. Therefore new ideas go through a stringent R&D and testing process that is designed to weed out products that the average consumer would dislike ('testing to the average'). The very logic that food companies use to inform and filter product innovation creates space for smaller players and entrepreneurs to play. And this is what lets the seed of disruption grow. Entrepreneurs who find a niche market for their unique food idea can be successful and do not have to worry about large companies stealing their share. If it so happens that the niche trait of the entrepreneur’s market starts gaining mass appeal, voila, ‘disruption,’ unless large players respond quickly enough either by launching their own products or through strategic acquisitions.

None of this is new, but even through food we can see that big players do not ‘miss’ new trends because they are clueless or conduct poor research. Rather, it’s a product of fundamental economics. In a resource constrained world, it is not rational for them to create products that appeal only to a subset of the market. One way to address the natural shortcoming that falls out from this logic is to integrate data, market analysis, and monitoring into the product innovation process. Continue to test to the averages, but as soon as your data team identifies a rapidly growing niche characteristic, act fast.

1099 Economy: Discussion Points

Generally speaking, we do not make enough time to discuss topics deeply and understand the fundamentals of the world around us. For that reason, I helped put together a group of people interested in exploring different topics through regular discussion. Our first conversation was about the 1099 economy. We did not (and likely will not) take proper notes, but I wanted to share a list of topics and questions we discussed because it was so interesting.

+ Leveraging Free Time - Is time 'free' out of desire or because no other opportunities exist? How will this change over time? As the economy improves? What is the real trade off between stability and flexibility?

+ Birth of the Uber-preneur / Flexibility vs. Exploitation of Labor that can’t get full time jobs with benefits

+ Regulatory risks that these companies face. From an investment perspective, how do you over come this? Adapting policy to match new technologies,

+ Benefits – who pays the burden? Is it fair to skirt the line and call your workers 1099s instead of W2s? As more people work as 1099s, does the burden fall on society if people are not covered (insurance), for example? How will W2 benefits change over time as a result?

+ Who represents the interests of 1099s? Unions across 1099 companies? Every man for himself?

+ Should Uber and other 1099 players be required to pay benefits if the time a worker spends on their platform exceeds a set amount of time (e.g. 20 hours)?       

+ Lack of loyalty in workforce and implications of maximal size of 1099 based companies

+ What skills are 1099s developing? Room for professional mobility. Do workers care?

+ What opportunities exist across all 1099 companies?

+ Does scale matter for these companies? Regional (yes). National (maybe for Uber, but not for most others)

+ Currently disrupting the service economy. Will it move into the knowledge economy soon?

+ Is automation the natural extension of  outsourcing to the 1099 market / abroad?

+ One of the reasons why firms exist is to reduce transaction costs between people. Technology changes this equation. Can this be extended to all divisions of a business? Can all functions be outsourced? What is the value of institutional memory? How can this be saved while outsourcing? When is it smart and not smart to outsource?