How to find the best INfluencers by area of expertise: An academic's guide

INTRODUCTION

LinkedIn is a social networking site with over 400 million members of whom at least 107 million are active. After Reid Hoffman launched it in May 2003, its Alexa rank of all webpages has climbed up to the 14th place. In deciding upon its name, the founder was inspired by Albert-László Barabasi's book from 2002 "Linked". Intrigued by this, I intuited a stronger intellectual grounding here than in another “social network” -that was born a year later for keeping in touch with family and “real” friends-, and became a member in ~2007.

REPUTATION RANKING IN A MEDIUM

Jeff Weiner as current CEO and Satya Nadella as CEO of Microsoft that bought LinkedIn less than a month ago (December 8th, 2016) aim at increasing engagement. Apart from networking for professional reasons, LinkedIn has become a media company and is characterized by a competitive reputation system. An important step in the process of LinkedIn taking on a media role was its recent roll out of the writing platform, and the "INfluencer" program that began on October 2nd, 2012. The current study focuses on the latter while being published on the former. The pervasive reputation/competition aspect is underlined by the possibility to compare one’s own profile views to that of one’s network (e.g. being in the top 3 or 5%) or in the total number of connections that reflects the adage of Lord Kelvin:

“I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind“.

However, a high circulation need not always be an asset, e.g. when the news is not evidence-based as discussed by Bob Sutton and Jeffrey Pfeffer already in 2006 (also cf. Cass Sunstein on "nudge" and Noam Chomsky on the "propaganda model"). Arianna Huffington wrote an article here about Fake news a fortnight ago and Jeff Weiner commented on the POTUS election in the immediate aftermath. These two articles were further inspiration for me, to develop a metric of quality (not unlike the "quality [arete]" of Bob Pirsig) along which to evaluate smoother, better written and better grounded from weaker, more superficial articles. This may not always be easy, but a lot is to be gained there with the PageRank demonstrating that I am not exaggerating. My personal experience is that profound talent in universities, while usually not remaining completely hidden, occasionally either gets shrugged off, "burned" or institutionalized. Cases include Shinichi Mochizuki, Grisha Perelman, John F. Nash Jr., Alan Turing, Kurt G?del (who may have predicted ~the Donald in 1947, cf. an upcoming post of mine), Ramanujan, Georg Cantor, Alexander Grothendieck as a few above average mathematicians and Robert M. Pirsig, with the latter two sympathetically having lived ~in the woods for decades respectively temporarily (relations to Heidegger, to the Luddites and to a Weltanschauung of "mindfulness" cannot be excluded with absolute certainty here). As those cases show, science/mathematics/writing significantly values conservatism. However, in spite of all its shortcomings, especially in challenging, skeptical times it remains for me the way towards progress and universalism par excellence.

BACKGROUND 

Another reason for the present study is that Pulse kept suggesting to me "INfluencers" with a six or seven digit number of followers who I found not quite deep. One hadn't bothered to look up wikipedia and called a candidate, a Nobel prize winner, another one invented a Shakespeare quote (while calling the bard a special friend of his), a third one wrote about Walter Mischel’s “marshmellow study” at the behest of the LinkedIn CEO without ever mentioning the original investigator, instead apparently wanting the kudos for himself. Some might say, "who cares?", but it is my conviction that sloppy thinking is at the heart of crucial historical events starting with "flash crashes" moving on to Farage- or Trump-like thinking. Within a few minutes of reading, I pointed out misattributions and when no reaction came realized I would be wasting my time with them as they were neither giving credit where credit was due nor relying on facts, but were in the game mostly for reasons of “ego” (cf. Goffman’s "presentation of self in everyday life" or Grant's notion of "takers" rather than "givers"). It has been commented that engagement on LinkedIn may be a general waste of time except for job hunt and resume. Actually, a Forbes article claimed “Jeff’s plans have since gone awry, apart from unmoderated and marketing spam in Groups, removing functionality from users without warning,

the publishing side has become an unprofessional parody of Facebook, awash with memes, maths puzzles, and pictures of Minions.”

However, I decline to give up on Linkedin, believing that the publishing platform can -with a few reasonable changes- be turned around into a solid project. Support comes from my realization that several INfluencers are hidden gems with the problem being mainly that I encountered them by "serendipity" only (cf. Robert Merton). For instance, after following Ian Bremmer a few days, commenting on a dozen of his posts, inviting him and us getting connected, via his connections I became aware of Frank Wu and the same process ensued. Of both professors I have had an impression of knowledgeable, smart, multi-faceted and warm persons, influencers in the best sense. More ought to be done to identify such individual sparks, and this is where the present study begins.

Personally, I had been using google news for a decade on a nearly daily basis. Actually, I read it in various languages with a focus on English, German, Roman languages in order to triangulate and reduce “bias” by comparing various "frames" (cf. behavioral economists such as Dan Ariely). Occasionally, I read articles on “project-syndicate” (also amazing for studying world languages), but found the comment section lacking engagement with the writers rather than merely with other readers. Academia is currently undergoing a huge transformation, with "early adopting" members realizing that just as privacy is becoming a hallucination, so is double-blind peer review. This is leading gradually to open access, even though for at least one millennial it may be coming at a singular cost (also check out Grisha Perelman and Shinichi Mochizuki who published online rather than through paid and peer-reviewed print publishing). To me the future is even more interactive and collaborative, Wikipedia-like, mirrored in the polymath project started by Fields medalist Timothy Gowers. The trend of interaction can be seen on LinkedIn as well, especially when articles are written by Influencers or capture the spirit of their times (the Hegel'ian Zeitgeist), again most likely due to reasons of Spence's and Akerlof's "signaling".

TARGET OF THE STUDY

In case

the present article leads members to find INfluencers that they want to follow, learn from and discuss with, or to reflect on which text really fulfills a purpose, thus "streamlining" whom to follow (aka Occam’s razor),

this article has served its purpose. Apart from this very applied goal, there is an academic goal beneath, namely to reach a better understanding what drives engagement and why do some articles and/or people become more of a "focal point" (Thomas Schelling) than others. Hypotheses that were measured and coded in the present study include “age” (classified into four demographic generations), nationality, language and work in academia based on job title. That age (and indirectly, generation) plays a role is an old and intuitive assumption in marketing and entrepreneurship where it is known as the “first mover advantage”. What is relevant here is that high earning celebrities who are young tend to do extremely well on Twitter in follower numbers (Katy Perry, Justin Bieber, Taylor Swift with 80+million) but not so much on Linkedin, where e.g. Lena Dunham had less than 200k [6], or Gwen Stefani around 50k [1]. Likewise, well-known politicians do not do well automatically (Sarkozy 7k [19], Jeb Bush 13k [1], Ted Cruz 20k [2], Obama 36k [1]) though some do (Ban Ki Moon 1,200k [62], Modi 1,700k [32] and Cameron 2,100k [108]), with myself putting 2016 year end number of posts in brackets. Obviously, it costs a certain amount of time and effort to raise engagement and gain people’s trust. Again, LinkedIn’s target is quite different from Twitter’s, since people can write LONG posts, and from Facebook’s, since one signals a follower choice to the workplace. For instance, a recent post defending the mind-blowing but often repeated suggestion of their difference got around 5,000 comments. On the other hand, given a person presiding over “apprentices”, there may be some convergence between the business world and show business, and given Artificial Intelligence encroaching on our private lives and lifting the barrier between both areas, it is also not clear that the two will be able to stay clearly “demarcated” (in the sense of K. R. Popper). However, what is clear is that to spell out an argument of a certain depth and complexity, space and hyperlinks are needed, as inclined readers will see in the present article, too.

Writing in a language understood by many, and living in a large country that for decades offered the world role models (James Stewart, John Wayne, "The karate kid", a manipulated "Marty McFly" scene), reduces transaction costs (cf. Oliver Williamson, as well as Herbert Simon on "transaction costs/search costs" and Robert Mundell on OCA). Gaining a doctorate and working in respectively being affiliated to academia reflects “signaling” to potential clients, coworkers and supervisors (or as we may here rightly say, to the "general public") certain skills and a mental resilience.

PROCEDURE OF THE STUDY

The author of the present study took a pdf source from May 11th, 2016 that listed all Influencers with picture, follower number and job title, as starting point. Afterwards, based on the pictures, gender was coded as male/female for all members of the set. Subsequently, a sample of the hundred INfluencer with the largest follower numbers was taken and for this sample, Wikipedia and other online resources were used to find out birthdates and nationality. In case this was not possible, I used an extrapolation by assuming that the year of graduation with Bachelor degree was the year the person turned 22. Such guesstimates are marked with a "circa" in the respective entry, and the author is grateful for corrections and additions, kindly asking to provide a reliable source if possible. Based on the sample, a correlation analysis was made between year of birth and follower numbers/follower rank. For all INfluencers, statistics were additionally gathered about “academia” (working as professor, dean, university president counted full, other titles were coded as “affiliated”) and about the language in which the person had chosen to describe his/her job title.

RESULTS

At the time of data collection, there were 752 INfluencers on LinkedIn. The average INfluencer has 234,000 followers, with a standard deviation of 573,000. While the maximum is 9.2 million, the minimum is 1,052, with the difference not being based on marginal productivity but on relative differences between individuals (Lazear and Rosen 1981, also cf. power law distributions).

AGE

Agewise, the average Top100 INfluencer turned 55 years last year with a standard deviation of 11.4 years assuming a Gauss distribution. Consequently, the range of plus one sigma to minus one sigma, in which there are statistically 68 %, is the age of 44 to 66; and the range of plus two sigma to minus two sigma in which there are statistically ca. 95%, is 33 to 78 years.

This means that the average INfluencer, being born around 1961 will be approaching in the next five to ten years the verge between baby boomer (1946-1964; 30 out of the Top50, 54 out of the Top100) and generation X (1965 bis ca. 1980; 15 out of the Top50, 30 out of the Top100).

If we go further to the outer margins of the Gauss distribution, both the younger generation of millennials and the older one of the silent generation represent each four to ten percent. Pete Cashmore and the first-placed academic, Adam Grant, are the two millennials in the top fifty; and Jack Welch, T. Boone Pickens and Ban-Ki Moon represent the silent generation (1945 and before), with the imbalance growing in the Top100 where there are nine from the former and five from the millennial generation (1981 and later).

A correlation analysis revealed that for the Top50 INfluencers, year of birth correlates -0.33 with follower numbers, and for the Top100 INfluencer, the correlation between the two is -0.23. This means that there is certain tendency that older people have more followers, a tendency that is more pronounced at the top. More technically, the variance explained by year of birth is 11.2% for the Top50 (the correlation squared), and 5.1% for the Top100.

GENDER

The gender representation was 3 males to 1 female. More precisely, 558 men make up 74.2% of the whole sample, while 194 women represent 25.8% (27% in the Top100). The numbers themselves are close to female representation at leading tech companies as well as faculty at top universities. Compared to the Nobel Prize, the Influencer percentage of females is much higher (48 women among 870 winners since 1901 with a growing percentage over the decades).

 NATIONALITY and LANGUAGE

As concerns both gender and age, INfluencers apparently reflect distributions in other global elite networks, apart from cases such as active elite sportspeople where (Post-)Millennials are presumably dominating. This is not the case concerning citizenship or language. In the Top100, three fourth have US citizenship (75%), 8% British citizenship and 19 nationalities share a percentage between 0.33% (Carlos Ghosn’s citizenship is divided into three parts, Brazilian, Lebanese and French) and 2%. For instance, concerning Nobel Prize Winners, Americans make up 363 of 1,076 or 34% (with there being at least one Nobel winner among the INfluencers, viz. Kailash Satyarthi), as concerns Olympic medals, they make up 2,802 out of 18,511 or 15%. Shares of sixty or more percent reflect rare samples such as worldwide top celebrities or best-earning sportspeople.

The main explanation I have for this phenomenon is that the underlying mechanism is bootstrapping. In other words, who becomes an INfluencer is strongly dependent on having other INfluencer’s supporting one’s bid (initially, one could apply, now the process is based on nominations). Given early INfluencers were mostly Americans, or wider part of the Anglosphere, this reflects what Lipman-Blumen and Moss Kanter called “homosociality”, even though here it is rather “homonationality” (also cf. “homophily”). In my research history, my main trademark, “The Panopticon Puzzle” was quite well received by American (and Anglo-Saxon in general) intellectuals, getting feedback by Chomsky, Pinker, Chalmers, Dennett and Ted Snyder, before getting a comment by Fields medalist Cédric Villani. The issue of the Melting pot is here relevant, as people who immigrated to the US traditionally gave up their former nationality.

While nationalities were encoded only for the Top100 INfluencers, the author categorized all “job title” descriptions (provided by INfluencers themselves) by their language. Here we see an interesting distribution. Up to the Top600 INfluencers, there are few people using a language other than English. Actually, in the Top100 (one Chinese), the Top200 (+1 French), the Top300 (+1 Brazilian), the Top500 (+2 Brazilian) and the Top600 (+1 Brazilian), it is exactly one percent. So while in the Top600, we have 1 title in Chinese, 1 in French, 4 in Brazilian (and ~594 in English), from then on it gets interesting. In the Top700, still 1 title is in Chinese, and 1 in German, but 7 are in French and 13 in Brazilian. Among all the 752 INfluencers' job titles, there is 1 in Chinese, 2 in German, 12 in French and 18 in Brazilian/Portuguese, with nearly 96% in English (LinkedIn currently has 24 languages available, meaning that 19 are not used by any of the INfluencers). Please be reminded that this is just the language in which the job description is given. There are many more Chinese, German and other speakers represented, but those chose to describe their job in English, without Hanzi or without German/French/Brazilian terms.

As an anecdote, follower numbers do not always mirror actual influence. Wladimir Klitschko has ~30.000 followers, while three other INfluencers who are 1st degree contacts of mine have six or eight times as many followers, but still got fewer profile views according to LinkedIn (see screenshot on top). Apparently, a long-time world championship title matters to members across the board, over and above those with sports affinity.

 THE SCIENTISTS/ACADEMICS

The author considers himself first and foremost a scientist rather than a private investigator, but identifying academics was really the hardest part in the present research project. Based on their job title, I discovered 24 “professors” (one had this word in French). After that, I found 4 deans followed by some presidents. And from then on, things got messy, including academic directors, chairs, fellows and others. Some are clearly professors but had e.g. put as there job title, “Imagine enjoying the exact flavors that you love... in less than a minute”. Dan Ariely, consider yourself identified ;). In total, the number of people within academia was fifty-six or 7.4% of all INfluencers. Of these, 38 are professors, deans, university presidents and the like, and 18 are researchers, “coaches” or directors affiliated to an institute of higher learning. In case you are interested in learning more about them, I sorted them by area of expertise, with women underlined.

•  MANAGEMENT (26): Adam Grant, Lucy Marcus, Joel Peterson, Sanyin Siang, Michael Wheeler, Anita Elberse, Jeff deGraff, Bob Sutton, Jonah Berger, Vivek Wadhwa, Michael Skok, Geoffrey Garrett, Steve Knight, Santiago Iniguez (the original spelling should be I?iguez, KK), Andrew McAfee, Yossi Sheffi, Michael Schrage, Ben Mangan, Dominique Turpin, Gérald Karsenti, Bill George, Jean-Michel Blanquer, Jeffrey Pfeffer, Wladimir Klitschko, Peter Todd, Laurence Capron,
• POLITICAL SCIENCE/POLICY/GOVERNMENT/ADMINISTRATION (7): Ian Bremmer, Bruce Katz, Robert Reich, Peter Sands, Dana Goldman, Janet Napolitano, Michael Crow
• LAW (7): Daniel Solove, Randy Kessler, Frank Wu, Erica Ariel Fox, Eric Goldman, Cass Sunstein, Ronaldo Lemos,
• ECONOMICS (3): William de Viljder, Dan Ariely, Tyler Cowen,
• MEDICINE (3): Lucien Engelen, Phyllis Wise, Michel Lejoyeux,
• PSYCHOLOGY (2): Chris Stout, Barry Schwartz,
• ENVIRONMENT/SUSTAINABILITY (2): Anthony Leiserowitz, Jeffrey Sachs,
• ROBOTICS/COMPUTER SCIENCE (2): Sebastian Thrun, Anant Agarwal,
• MEDIA (2): Joichi Ito, Tom Bedecarré,
• COGNITIVE SCIENCE: Don Norman,
• PHILANTHROPY: Lucy Bernholz.

Such an overview may also be helpful to the INfluencers themselves, as soon as they start realizing that they are not competing against each other, but furthering a common conversation with the goal of finding out how best to tackle global challenges.

I also investigated the complete list of institutions with which the 56 Influencers are affiliated, counting 42 (75%) inside the USA and 14 (25%) in other countries. Within the USA, the leading affiliations are Harvard (7), Stanford (7), MIT (4), UPenn (3, all at Wharton), UC Berkeley (3 incl. one at Haas), Duke (2, both at Fuqua), and with one each come Yale, Columbia University, UMich (Ross), Emory University, Swarthmore College, U Illinois in Urbana-Champaign, U Illinois in Chicago, George Washington University, George Mason University, NYU, UC San Diego, USC, UC Hastings, Santa Clara University, Arizona State University and the Brookings Institution. Outside the USA, France dominates with 6 out of the 14 which include: INSEAD (2), HEC (2), IE (2), ESSEC (1), Université Denis Diderot (1), St. Gallen (1), IMD (1), Radboud University (1), Ghent University (1), Imperial College (1) and Rio de Janeiro State University (1) as the only non-European one. While the number of 75% is very close to the number of US nationals in the Top100, the remaining 25% have a significantly stronger link to continental Europe than the citizenship.

LIMITATIONS OF THE STUDY

Even though usually available, on this occasion I did not code any information about either ethnic group, industry, city of residence or number of long posts. Also, years of birth and nationalities were limited to the Top100 of the list. Furthermore, the follower numbers may have evolved somewhat in recent months. These limitations leave place for further study.

Renowned sociology professor Mark Mizruchi (University of Michigan) did a well-known study about board interlocks. So given that INfluencers accept the author’s invitation, one may look here at “INfluencer interlocks” and to what extent it correlates with follower numbers. Other criteria may also be interesting, and I strongly encourage readers of my articles to offer their feedback in the comment section.

SUMMARY: TL;DR

Working and breathing in the marketplace for ideas called academia, I have trained my sensors for self-promotion and bullshit on a daily basis for over a decade. Also helpful was my work for one of the top four strategy consultancy companies in the marketing department and at one of the leading management schools. It thus contributed to identifying people with a strong track-record whose precise abilities I usually evaluate over the course of months, years or decades, as opposed to others who may be less civilized or simply less able. The present study explicitly listed academic INfluencers by area of expertise, finding unsurprisingly that over 70% (namely 40 out of 56) come from management (26), law (7) or administration/policy/political science (7). It also pointed out members of the two not so common generations who stand out (silent generation, millennials) among hundreds of baby boomers and Gen X'ers, as well as people who stand out for providing self-descriptions in a language other than English. A table of all INfluencers of aforementioned date and source is available from the author as another long post. Wishing a

happy new year 2017 to everyone!