Thoughts on informational velocity and the near-future of twitter. With references to @Oreo and @NNTaleb

On Optionalities[i] and Perpetual Betas[ii]:

Twitter and ‘Bursts of inconsequential information[iii] 

            Twitter is fascinating for many reasons. When you think about it, its founders – Jack Dorsey, Evan Williams and Biz Stone – have one true perpetual beta. It started with traditional blogs through blogger, which they sold to Google. Then they moved to podcasting with Odeo. Then you had twitter. Shortly after, they added integrated photo sharing. They acquired Posterous last year, and rolled out Vine early this year.

Long-form, audio, short-form, pictures, long-form, short-video. These are fundamentally all experiments with form and style in communication. Where Facebook is the summation of yours, your friends’ and families’ thoughts and experiences, twitter is more like Google – a vast open mindscape that forces the communicator to likewise condense both the profound and the banal into 140 characters.

Twitter’s business model, however, can’t mimic either – at least, not in its current design avatar. Facebook is where you go to see what your friends are up to. Google is where you go when you need to find something specific, or general information about something. What is the behavior that twitter elicits? One might argue that twitter is where you go to have a conversation. A river you dip in and out of from time to time. We know a river is ever-changing, but twitter has not only speed but also informational velocity. Without any form of algorithm to slow down the feed, there is no flavor of the month, week, or even day. The informational velocity on twitter can change by the hour or minute.

Now I would argue that what advertisers – the business-model of choice in social – pay for is fundamentally predictability. If you look at the ad metrics of old you will see that the reason ‘soap operas’ exist is because TV dramas with a high probability of housewives watching them became the preeminent contact point for FMCG firms selling soap. So how does that work in a system that thrives on randomness and chaos? Well, maybe we can take pointers from two unlikely sources – Nassim Nicholas Taleb – Professor of Risk Engineering at NYU, and, umm, Oreo.

Taleb’s life’s work has centered on the principle of philo-stochasticity – or, that which gains from chaos – which he today calls ‘the antifragile’[iv]. How can twitter gain from chaos? Well, in my opinion, creating a deeper ‘search’ function within the network (currently tweets disappear from search after a certain point) may allow for twitter to better position its promoted tweets to maximize clickthroughs – they already do this with trending topics – but this is not quite it.

What I believe they can add is what we’ve learned from Oreo’s hijacking of the superbowl blackout[v] on twitter: create a two-speed advertising network – one that allows advertisers to promote tweets/promos around topics, and one that gives advertisers data for changes in informational velocity that can help them make more responsive campaigns harnessing relevance in current events. Trending #tags do this somewhat for users, but it needs to be better visualized, and flagged earlier to be of use to marketers. That may well be where a successful future of twitter lies. Volatile. Adaptive. Philo-stochastic.

Endnotes and References


[i]The value of additional optional investment opportunities available only after having made an initial investment.’  

[ii] The open source dictum, “release early and release often” in fact has morphed into an even more radical position, “the perpetual beta,” in which the product is developed in the open, with new features slipstreamed in on a monthly, weekly, or even daily basis.

[iii]Twitter creator Jack Dorsey illuminates the site’s founding document’


[iv] ‘Taleb on Antifragility’ [youtube]


Bigelow’s 14 Maxims for Ideal Prognosticators, from George Dyson’s Turing’s Cathedral


An excerpt:

Bigelow compiled a list of fourteen “Maxims for Ideal Prognosticators” starting with MIP 1: “Make all observations in same coordinate system as will finally be used by the gun-pointer.”

Maxims 2–4 advised separating the available information into that needed immediately and that needed later, while Maxim 5 added that “if noise is ever to be filtered from signal, it must be done at the earliest possible stage rather than after the two are tangled with other noises and signals, for the same reason that repeater stations are used on a signal line rather than filters and amplifiers at the ends.”

Maxim 7 advised “Never estimate what may be accurately computed”; Maxim 8 advised “Never guess what may be estimated”; and, if a guess was absolutely necessary, “Never guess blindly” was Maxim 9.

Maxims 10 through 14 specified how to implement optimal prediction when the target “has the character of a Brownian motion impressed upon a resonator system.” Existing methods of tracking a target’s changing position “of necessity refers it to an irrelevant point of observation thus destroying its fundamental symmetry,” while an ideal predictor should assume that the target obeys the conservation laws of physics “upon which is superimposed a random modulation symmetrical in time.”

The Wiener-Bigelow debomber would model the behavior of the airplane within the frame of reference belonging to the airplane, rather than referring it to that of the observer on the ground.

Dyson, George (2012-03-06). Turing’s Cathedral: The Origins of the Digital Universe (pp. 112-113). Random House, Inc.. Kindle Edition.

Thoughts on Facebook’s Graph Search, and the (near) future of Social

I have had a few weeks to play with Facebook’s Graph Search since it was introduced as a Limited Beta earlier this year. It doesn’t (yet) do a lot of what I expected of it, but it has been interesting and, in some ways, ‘surprisingly’ useful.

Why use the word surprising? Because of the relational nature of how it works. At present, it works at the object level of people and pages. What does that mean? – You can use it to discover links between people, or between people and their photographs, places, restaurants, games, music, movies – but you can do it only at the ‘page’ level.

Having ‘liked’ a page adds it to the pool of data – but specific posts, or “likes” on the broader web through Facebook’s Social Plugins don’t yet reflect within search parameters. Graph Search presently shows us some new ways of connecting people – through status or interests – that didn’t exist earlier.

But let’s face it. What we’re really waiting for is the next phase – search access across the entire Graph. Why is that interesting? Let’s look at it from three different perspectives:

A)   Users: For active users, Timelines serve as a fantastic repository of data that we have discovered or shared during our time on the network. For the pool of idea sharers on Facebook, enabling search at this object level would truly make search interesting – it certainly beats 30 scrolls to get to that report you posted a month ago.

In terms of networks, Twitter is a river – you don’t necessarily expect to go back and get data – but it still offers a ‘Favorite’ button – which is a lot like bookmarking content you liked. Facebook on the other hand, is the sea. A more permanent repository of your mindspace, as close to a ‘quantified’ virtual self as you can get. You could even make the argument that services like Second Life never quite took off, because their roles were subsumed by something more organic and less alien to behavioral adoption – Facebook.

Parsing Facebook’s Timelines real-time would potentially collapse their servers today, which is probably why they aren’t yet there. But as their Graph Search algorithm evolves, and server processing horsepower catches up, this will be an incredibly important tool for users to manage their own data / interact better with data their friends post to the network.

B)    Monetization: One might argue that Facebook’s big monetization challenge is harnessing intent. I argued in a short essay a couple of years ago that Google didn’t really need Google+ to take off as a network as long as people populated their “+1” database. When users search on Google, they are, by virtue of keying in their search query, expressing intent – and since Google’s monetization comes by being relevant in search, all they really need is a +1, which is almost as good as a proper “post” in terms of indicating endorsement by the user. Posts help in creating stronger endorsements – so posts with more +1s might be considered to be more relevant, or ‘cleaner’.

Google’s ‘graph’ is called the ‘Knowledge Graph’ – because its job is to give you data you want. Facebook’s graph today is called the ‘Open Graph’ – but what it really needs to develop over the next year, is an ‘Intent Graph’. Intent. That is what you need to figure out when monetizing through advertising. Google has intent because you’ve told it what you want. Facebook needs to uncover intent, understand latent needs, trigger conspicuous forms of consumption (forms of social buying outside of the sort of discounting practices we’ve seen on the web so far, more like events, concerts, even utilities – but more on this in “C”)

Broadening Graph Search to smaller objects will very likely take a lot more processing horsepower as I mentioned earlier, but it is at that granular level where Facebook will be able to move from serving up ads that you might be interested in, to projecting what you want. Again, this is not meant to be intrusive or glorify ‘conspicuous consumption’, but rather to help certain forms of real-life behavior translate more organically to the social web.

What is it that we buy? Advertisers buy ad time because of metrics that offer them a certain “predictability” linking ad spend to purchases. Consumers use products or services because they offer them a “predictable” outcome through consumption. That is the purpose of the notion of the ‘intent graph’ – to make behavior and consumption predictable, and therefore scalable and sustainable. Which leads me to C.

C)   The Invisible Continent: We’ve seen the stories about Facebook being larger than most countries in the World in terms of active users. What’s next? Well consider three ‘macro’ factors we’re looking at around the world today:

a.      There are pages and pages of literature now on subjects like this, and a broader cultural movement that may or may not be related to the demographic shift towards Gen’s X, Y and Z that trust conglomerates less and demand more individual growth and fulfillment.

b.     We’ve seen the rise of artisanal endeavors, micro-manufacturing, and re-localization of businesses; alongside the rise of technologies like 3D printers that further facilitate / accelerate this change. And

c.      We’re seeing a broader breakdown of the old economy. Governments are struggling to resuscitate growth because of many reasons. Confidence and liquidity are two sides of the same coin, and there isn’t as much confidence as there used to be, even with low interest rates in many economies alongside tax breaks to entrepreneurial endeavors. And finally, mega corporations are struggling at two levels – growth, and finding the right people. And while there are calls to ‘fix education’ to produce more skilled labor to work for corporations, what’s more likely to happen is further contraction for major corporations and the rise of smaller businesses.

There are four firms positioned to help birth this new World today. The two that are most ready are amazon and eBay. Their respective ‘Marketplace Seller’ models are a viable backbone for this kind of behavior. And then there’s Google and Facebook. What Facebook has that the other three don’t, is a ‘social’ environment with an Apple-esque ‘Walled Garden’ in terms of design and function. They can facilitate while still controlling aspects of design, alongside limiting potential downsides like oversharing / commerce featuring explicitly on your newsfeed – something people are unlikely to want. Facebook could do for brick-and-mortar startups – what Apple did for app developers – create an economic ecosystem overnight. Sure, this point is further out than the more immediate natures of A) and B), but this is perhaps something to watch out for.

This in itself would have a whole new set of questions raised. Payment gateways, fulfillment, jurisdictions in the immediate, and broader potential questions of taxes, levies, unified currencies in the future.

Point C is, of course, a Facebook-centered look at the future of social, but one which is entirely plausible over the next decade. Whether it is Facebook or another firm that does it, it seems for now that the winds of culture are blowing this way.


Good read -> The Best-Performing CEOs in the World – Harvard Business Review

It’s no accident that chief executives so often focus on short-term financial results at the expense of longer-term performance. They have every incentive to do so. If they don’t make their quarterly or annual numbers, their compensation drops and their jobs are in jeopardy. Stock analysts, shareholders, and often their own boards judge them harshly if they miss near-term goals. And without equally strong pressure to manage for a future that stretches beyond 90 or 180 days, CEOs’ behavior is unlikely to change. Developing a simple yet rigorous way to gauge long-term performance is crucial; after all, in business, leaders default to managing what’s measured.

Five years ago we launched a global project to address that challenge. But we wanted to do more than just devise the right metrics. Our goal was to implement a scorecard that would not only get people talking about long-term performance but also alter the way that boards, executives, consultants, and management scholars thought about and assessed CEOs. We wanted this innovation to shine a spotlight on the CEOs worldwide who had created long-term value for their companies, and we wanted to give executives around the world critical benchmarks they could aim for.

Three years ago, in the January–February 2010 issue of HBR, we introduced such a scorecard. It evaluated chief executives on their entire tenure in office. We used it to rank the performance of nearly 2,000 CEOs. This month we are publishing a new version of that analysis. We have expanded it along two important new dimensions—making the group of CEOs we studied truly global, and examining which CEOs and companies were able to do well not only financially but also in terms of corporate social performance.

Judging CEO Performance

For the most part, we used the same methodology that we did three years ago. (See the sidebar “How We Created the Scorecard.”) We wanted to accomplish three things:

Assess the long-term performance of each CEO, from the first day on the job to the last. (Or for CEOs still in office, until August 31, 2012, our last day of data collection.) To do this, we looked at how much total shareholder returns had changed over that time period (adjusting for country and industry effects), plus the overall increase in market capitalization.

Reflect the global nature of business. In 2010 we drew candidates from the S&P Global 1200 and BRIC 40 lists; this year we worked with three other emerging-market indexes as well. The pool of CEOs studied increased by roughly one-third, from 1,999 in 2010 to 3,143 this year.

Be objective. Other rankings use reputation and surveys, which tap into popularity and celebrity status, to score CEOs. Instead, we use only performance data—notably, total shareholder performance. Other metrics, such as sales, profitability, and innovation rates, are useful, too, but they differ by industry, which makes comparisons difficult.

How We Created the Scorecard

Granted, one downside of using truly objective measures is that our ranking may not exclude CEOs who have disappointed stakeholders on dimensions where performance is more subjective. This can be especially challenging with CEOs from emerging markets where rules are still being established. But even though CEOs are held accountable in areas that shareholder performance cannot capture, it remains the principal standard by which they are judged.

Who’s Up, Who’s Down

The top 100 CEOs on our list performed exceptionally well. On average, they delivered a total shareholder return of 1,385% during their tenures and increased their firms’ market value by $40.2 billion (adjusted for inflation, dividends, share repurchases, and share issues). The contrast between their results and those of the bottom 100 CEOs was striking: On average, the bottom 100 produced a total shareholder return of -57% and presided over a loss of $13.6 billion in market value.

It comes as no surprise that the best-performing CEO over the past 17 years was Steve Jobs of Apple, who was #1 on our 2010 list as well. From 1997 to 2011, Apple’s market value increased by $359 billion, and its shareholder return experienced average compound annual growth of 35%. That remarkable accomplishment is likely to go unbeaten for a long time.

Jeff Bezos of has now climbed to the #2 spot, up from #7 in our 2010 list. Under his leadership, the company delivered industry-adjusted shareholder returns of 12,266% and saw its value increase by $111 billion. In recent years the online retailer has expanded aggressively into new segments such as cloud-based computing services, while working to get the most out of the markets it already occupies. Its revenue growth shows no signs of slowing: Sales increased by 40% in 2011.

“You Have to Be Willing to Be Misunderstood” 

The highest-ranked woman on the list is Meg Whitman, currently the CEO of beleaguered HP, whose performance as the CEO of eBay from 1998 to 2008 earned her the #9 spot. Overall, only 1.9% of all the CEOs we studied were women.

Top 5 Female CEOs

There was considerable turnover in the top 100—this year about half (43) of the CEOs are new to that list since 2010. The change happened largely because in creating the 2013 ranking, we examined a bigger, more international sample of CEOs. To make the top 100 in 2010, CEOs had to rank in the top 5% of the executives in the study; this year they had to land in the top 3.2%. The majority of newcomers to the top 100 are from emerging markets, not surprisingly, but eight U.S. CEOs made the cut for the first time.

One notable new name is Lars Sørensen, the CEO of Novo Nordisk, a company that made its name selling insulin for diabetics. He shot up from #233 in 2010 to #20 in our current ranking. After Sørensen took over the company, in 2000, he spent 10 years and $500 million expanding the sales force to make it truly global. When all his competitors were investing in diabetes pills, Sørensen shut down Novo’s pill research and instead focused the company on its core competency—insulin and other injected diabetes medications, notably prefilled insulin “pens” that eliminate the hassle of using a vial and syringe. His bet that sales would continue to grow in the wake of a worldwide diabetes epidemic has clearly paid off.

Top 5 CEOs Europe

Even industries that have gone through tough times in recent years have seen some exceptional results. Airlines, for example, have not been doing well, but Air China, under the leadership of Li Jiaxiang (#17 on our list), bucked that trend. During his tenure, from 2004 to 2008, the company’s total shareholder return was 1,022 percentage points higher than the average for its industry peers, while its market capitalization rose by $37 billion. With his guidance, Air China obtained a 50% share of the market in major Beijing airports and joined the largest airline alliance in the world.

The World of CEOs Is Not Flat

With a truly global sample to study, we can do a better job of comparing countries and regions. This is a significant advance; for decades most analysis of CEO performance has been U.S.-centric. Now we’re able to examine data country by country—and we find real differences when we do.

China has been the growth miracle of the past decade, so you might expect CEOs there to have done very well. We find that the opposite is true: Among the 3,143 CEOs we analyzed, the average rank of Chinese executives was 176 places lower than the average rank of U.S. executives. Only three Chinese companies’ CEOs made the top 100, though 17% of all the executives studied were from China. The Chinese leaders we asked about this discrepancy theorized that as the country’s companies become more innovation-focused, their performance will improve.

Top 5 CEOs China

Likewise, the average rank of Japanese CEOs was 562 places lower than that of their U.S. counterparts, although this is not a big surprise, since Japan’s economy has struggled for many years. On the whole, U.S. CEOs did not shine either, despite holding six of the top 10 slots. Their average rank was 215 places lower than Latin American CEOs’, 140 places lower than Indian CEOs’, and 137 places lower than British CEOs’. Continental European and U.S. CEOs ranked about the same. U.S. CEOs have not been as competitive on a global scale as one might think.

One bright spot is Brazil, whose CEOs make up only 4.5% of the total sample but 9% of the top 100. They include Roger Agnelli of Vale (#4) and Embraer’s Maurício Botelho (#11). Botelho took over the state-owned company in 1995, when it was reporting losses of around $300 million a year, and over the next 12 years, built it into a world-class competitor. (Interestingly, Brazil is also overrepresented in the bottom 100, suggesting that companies from that country play a high-risk, high-reward game.) Another standout is Mexico, whose average CEO ranked 108 places higher than the average U.S. CEO.

Top 5 CEOs Latin America

The national and regional differences get even more interesting as we turn to the question of why some CEOs got better results than others.

What Accounts for Success?

Our 2010 article looked at several factors that might be relevant to good performance (whether CEOs were hired from inside the company, had an MBA, and so on). We tracked those factors again, and our global comparison revealed some insights into differences across the world.

The insider-CEO story. Management thinkers have long debated whether it is better to appoint an insider as CEO or get someone from the outside to run the company. But most studies have focused on U.S. corporations.

In our full sample of 3,143 CEOs, 74% were insiders. India had the lowest proportion (63%) and Japan the highest (90%). Overall, insiders did better than outsiders; the insiders’ average rank was 154 places higher than the outsiders’. This is similar to what we found in the 2010 global ranking. It didn’t hold true, however, in major parts of the world. Insiders got better results in the United States, the United Kingdom, and Latin America, but there was no difference between insiders and outsiders in continental Europe, China, and India.

Top 5 CEOs India

What about the idea that outsiders are preferable when a company is in trouble? We find that boards—especially in the United States and Europe—do have a slightly greater tendency than normal to hire outsiders when the company is underperforming (measured as having an industry-adjusted total shareholder return of -24% or worse for the two years before the CEO started). But the results those outsiders produced varied by region. In the United States they didn’t get better performance from struggling companies than insiders did. In Europe outsiders did better; the average rank of those who took over subpar performers was 370 places higher than the average rank of their insider counterparts.

In Latin America, however, the picture was different: The average rank of insiders who had taken the helm of poor performers was 750 places higher than that of outsiders who had. Regional factors help explain this disparity. A large number of Central and South American firms are family controlled; another large segment is government controlled. Business families—and, in some cases, governments—exercise a strong influence on long-term strategies and investment decisions, which makes it more difficult for a CEO who is new to a company to operate.

The upshot: In the United States, outsider CEOs usually do not deliver the goods, whether the company is underperforming or not. But this finding can’t be generalized to other parts of the world. Boards need to keep regional success factors firmly in mind when selecting CEOs.

The curse of great prior performance. If you want to create a lot of shareholder value, it pays to take over a company that hasn’t been doing well—at least if you’re in the United States, China, India, or the United Kingdom. In those countries a poorly performing predecessor is often followed by a high-performing one. But there is no such effect in continental Europe, Japan, and Latin America.

The Legacy Litmus Test

The greater continuity in company performance in Latin America is probably a reflection of the long-term control exercised by business families, investor syndicates, and governments, whose visions don’t change even as CEOs come and go. In Latin America those parties generally make the important bet-the-company decisions and policies, while CEOs are mainly responsible for execution.

An MBA degree. In the wake of the financial crisis, MBAs were accused of being value destroyers. We supplied the debate with some contrary data in 2010, showing that the average MBA ranked 40 places higher in the study sample than the average non-MBA. We saw similar results in this year’s list. In this case, we did not discover that CEOs of certain nationalities benefited more from an MBA than others.

Industry. In general, industry differences count for very little in our analysis, explaining only about 5% of the variation in CEO performance. That said, the high-tech industry is overrepresented in the top 10, with five CEOs, including’s. CEOs from the energy industry, who make up only 5% of the group studied, are overrepresented in the top 100, where they occupy 15% of the slots.

Doing Well and Doing Good

Many management thinkers argue that it is no longer enough to do well financially; companies also need to improve the well-being of (or at least not harm) the communities in which they operate, the environment, and their employees. (See, for example, “Creating Shared Value,” by Michael E. Porter and Mark R. Kramer, HBR January–February 2011.) That’s the good news. The bad news is that stellar performance on both dimensions is no common or easy feat.

This year we examined the correlation between the financial performance of leaders on our list and their social and environmental performance as measured by MSCI, a highly reputable firm that rates major companies. Despite all the rhetoric, we discovered that the correlation between the two sets of data is, well, zero. You can see this clearly in the exhibit “Does Doing Good Help CEOs Do Well?” Companies are scattered all over this chart. Though many articles suggest that responsible corporate behavior—say, in sustainability—will automatically improve your bottom line, clearly it’s not as simple as that. Some companies probably aren’t managing with such issues in mind. Some may not have attractive social or environmental strategies; some may have misalignment between those strategies and the overall corporate strategy; and some may have incomplete measures of social or environmental practices.

Does Doing Good Help CEOs Do Well?

But the chart did reveal outliers. Five percent of the CEOs for which we had sufficient data fell into the box at the top right; they delivered great financial performance year over year and performed strongly on social and environmental dimensions. It is a rare achievement, indeed, but it is possible.

These trendsetting CEOs are the new role models for leaders pursuing the paradigm of creating shared value. One example: Franck Riboud of Danone, a French multinational with $27 billion in annual sales. Danone’s excellent financial performance earned him a spot in the top 10% of this year’s sample (a truly amazing achievement for a consumer goods company); at the same time, the company received extremely high ratings from MSCI. Another outlier is Natura’s Alessandro Carlucci (who made the top 6% for financial performance), a leader among CEOs who believe that alleviating poverty and inequality and protecting the environment are intimately tied to their business agendas. Carlucci and Riboud have both confronted the key social or environmental issue in their industry (in Danone’s case, obesity and unhealthful food consumption; in Natura’s, deforestation and poverty) and redirected their company’s strategy to tackle it.

We also looked at CEOs whose companies had high social and environmental performance in 2010 but whose financial performance kept them out of the top 15% of the group studied that year. Since doing both well and good can be a long-term strategy, we wanted to see whether any of those CEOs had then moved into the top 15% of the current financial ranking. We found four: the leaders of Adidas, Inditex, Hermès International, and Eaton.

At Adidas, CEO Herbert Hainer oversaw the implementation of a triple-bottom-line philosophy, a massive push to slash the company’s carbon footprint, and the increased use of recycled polyester as well as sustainably farmed cotton in products. One of Adidas’s latest sustainable innovations is DryDye technology, which removes the need for water in the dyeing process. At Eaton, Alexander Cutler has embedded sustainability into the company’s culture and practices. The diversified power management company develops innovative products and processes, such as hybrid electric and hydraulic power trains and electric power control systems, that help customers and consumers conserve resources and reduce their carbon footprint.

This new breed of leaders not only rejects the idea that financial market demands are more important than stakeholders’ needs but also demonstrates that companies can excel at meeting both. These CEOs have shown the way, and others can learn from them.

We don’t foresee a time in the near future when measures of social performance will be as objective as the measure of long-term financial performance we’ve developed. That said, we will continue to track how CEOs are doing in the two areas, with the aim of encouraging leaders to shine in both.

Everyone in the business world seems to agree that executives should be less obsessed with quarterly earnings and more focused on the long term—everyone, that is, except the decision makers who hire and fire executives and the people who buy and sell company stock. The short-term emphasis won’t change until a new paradigm for evaluating performance emerges. Talk alone won’t bring about that change; we also need a whole new method of evaluating CEOs. Here, we’re proposing two key improvements: a robust, objective measure of leaders’ performance over their full terms in office, benchmarking all chief executives of major global companies; and an assessment of the correlation between a firm’s financial results and its environmental and social practices. We hope that boards of directors, pension funds, hedge funds, and other shareholder activists will use these measures to better evaluate CEOs and to guide the selection of tomorrow’s leaders.