Too-much-information (TMI) syndrome is a chronic affliction of contemporary existence. The affliction is growing, without an app that can alert us to its threats and update us on the speed with which the disease of big data is invading ever more of our lives. Other IT buzzwords with mystifying potentials are artificial intelligence (AI) and fintech, short for financial technology.
For corporations, the disease of TMI takes the form of an addiction. Everyone seems to need big data, without necessarily having found a way to understand its specificities and implications. But there is a very human coping strategy that says, “If you can’t beat them, join them.” In local entrepreneurship, this maxim manifests itself through startups that catch on to external trends and produce a locally designed version of an app that is expected to benefit from the trend—in this case, the desire to analyze and benefit from big data.
Technical challenges related to big data give rise to extensive research needs and very costly development efforts for data storage and processing solutions on the scale of petabytes and exabytes, the latter term having recently been added to our vocabulary in order to describe a quintillion of bytes. Quintillion is what comes after million, billion, trillion, and quadrillion. Clear? (Not to this writer, sorry). Blessed are those who understand the mathematics needed for managing such numbers, but the business of capturing and managing these information volumes seems to be cut out for larger corporate entities rather than the typical two to four-person startup.
Startups that use big data in Lebanon wisely seem to concern themselves mainly with the more recent interpretation of the term—analyzing previously inaccessible data volumes for predictive analytics of customer behaviors to improve corporate performance in customer responsiveness. As such, big data apps represent the digital workshop of attention merchants and marketing magicians who nudge people to behave in ways that are most profitable to the company they serve, whether it is a manufacturer, distributor, specialized retailer, or large provider of commercial healthcare, insurance, or finance.
Restraints
Besides the requirements for large capacities on the tech side of big data, what also restrains the space for specialized Lebanese startups in the areas of fintech, big data, AI, and Blockchain or cryptocurrencies, are cultural and legal factors such as the conservatism of the financial industry and regulators. In conferences that promote Lebanese investment and economic potential, such as the Lebanese Diaspora Energy (LDE) event in late spring of 2017, fintech was touted as promising field of entrepreneurship, but the numbers of successful independent startups in this area are limited. Digital currencies were—until recently—officially disavowed by the central bank, and enabling distributed-ledger technology (Blockchain) apparently means that the ecosystem has yet to produce startups with a track record of operations in this realm.
Despite this local reality, there is absolutely no reason to question findings by global consulting groups that show a huge boost in fintech investments from Q1 to Q2 in 2017, reaching a total of $12 billion during that period—$8.6 billion of that was in Q2 alone, according to KPMG, says a December 2017 note by Arabnet. The note also quoted another report, by CB Insights, saying that venture capital (VC) investments in fintech in Q2 of 2017 reached $5.2 billion globally.
The numbers for MENA are next to nothing in comparison. Arabnet cites “over $24 million” in fintech investments in 2017, and refers to just three investment deals that are supposedly fintech-related. They actually add up to $24.96 million, with the largest of these investments being a $20 million injection by undisclosed parties into a three-year-old payment gateway called PayTabs that in an earlier round got funding from Aramco Investment Ventures. The other two fintech startups cited with funding in 2017 are based in Dubai. It is hard to imagine how MENA fintech investments might even have reached 1 percent of global fintech funding over the years, as Arabnet suggested.
When endeavoring to profile startups in areas such as AI, fintech, and big data, Executive was treated to a number of incomplete leads, rumors, and less-than-coherent startup narratives that suggested deficiencies in the communication strategies of these entities and their partners in the ecosystem. Also, this magazine has yet to see impressive numbers of new startups relating to AI or fintech that are not iterations of companies that we profiled in one of our previous annual reviews on Lebanese entrepreneurs—but Executive has to admit that its investigative capacity has been recently somewhat impaired because of human resource attrition. (A core editorial team member had to invest immense personal efforts into a near-term repatriation to the land of his birth; Matt, thank you for years of good work at Lebanon’s discerning voice of business and entrepreneurship!)
A big data startup sampler
Thus, in this 2017 roundup of entrepreneurship companies, Executive offers profiles of only two startups that are active in the space of big data retail analytics. The companies share a specialization in data collection from in-store environments for the purpose of predictive analytics and marketing optimization. What is striking about their differences, however, is that they represent diametrically opposite professional backgrounds in the small field of big data retail analytics entrepreneurs.
One company, Vision in Motion (ViM), is the brainchild of a fresh entrepreneur with no personal experience managing a fast-moving consumer goods (FMCG) enterprise. In fact, having started the company while still working toward his Lebanese high school baccalaureate, 19-year old Samy Khoury embarked on his venture with no enterprise management experience whatsoever.
He was an achiever that had been recognized in 2015 as young innovator at an international competition in Warsaw, Poland, but his entry into entrepreneurship came in the form of a lucky break. To join the Speed Beirut Digital District (BDD) acceleration program in 2016, he was required to have “a technical co-founder,” Khoury tells Executive. Not having found the required partner until the day before the application deadline, Khoury called a friend who owns a small grocery in Ashrafieh. As Khoury tells his startup tale, the friend answered, “I don’t know any programmers,” but a customer in her store said, “Yes, you know a programmer. He’s your friend from school.” This chance connection à la Libanaise led to a phone call between Khoury and this programmer, who established their partnership in the last minute to qualify for joining the Speed program.
The other company, eQuality, tilts to the other extreme in terms of the typical startup founder’s age and professional track record. Nadim Tabet incorporated the company in his mid-50s, and already had 30 years of experience in the FMCG game, 16 of them as a management consultant. “I’m a management consultant by profession, and [I] specialized in the FMCG business, where I worked with Procter & Gamble and managed companies that relate to the FMCG business,” he tells Executive.
His professional journey to set up eQuality included stations in the United States and Dubai before establishing himself in Lebanon. Here, Tabet met his co-founder, a computer engineer named Rafic Hage, through a work relationship. Hage, 37 years old and seasoned as founder of several IT startups, explains how he developed his passion to serve an industry with lack of technology in specific niche areas.
“I was attracted to FMCG because it was new to me. I have been in IT ever since I graduated and have started three companies, which means I was always in this scope of responsibility such as finding resources, hiring resources, setting up plans, scoping projects, and leading developments from scoping to delivery. What was interesting in the partnership with Nadim in particular was to do IT solutions for businesses within a specific vertical,” Hage tells Executive.
`ViM’s path in entrepreneurship last year involved three months of participation in the Speed@BDD program, and one month immersed in Silicon Valley upon invitation through Speed. The accelerator injected $30,000 cash and provided other benefits in exchange for a 10 percent equity stake in ViM. This year, the team won a number of competitions at Arabnet in Beirut and Dubai, coming to an agreement there with PepsiCo under which ViM will implement a test project in one or two Dubai supermarkets from January 2018, with an eye to expanding the relationship if this test goes well.
eQuality progressed on its own financial power and did not participate in any incubation or acceleration program. Tabet says that the startup is oriented to the sophisticated end of the market, and its trilingual app comes with upmarket flair when compared with competitors. He is exploring the use of image-recognition software for future applications, but describes the technology in the market as currently too costly and not yet mature. Hage says eQuality is geared toward using artificial intelligence: “In any project involving AI, you start with collecting data. You need a large set of information for you to mine,” he explains. “For the past two-and-a-half years, our customers have been collecting data, and now we have enough data to start coming up with the proper analytics and predictive analysis.”
According to Tabet, since eQuality entered business with a target of serving large corporate customers, 25 client companies in the FMCG space signed on for its first product, a “merchandising intelligence application.” Under the name eye2, it offers FMCG distributors and suppliers a new way to assess their products’ merchandising status inside hypermarket store environments. While data is entered through interconnected devices, processed in the cloud, and immediately displayed on a dashboard, data-collection and entry methods are not the core strength that eQuality is focused on. Rather, the type of data collected and its fast translation into percentage-share visuals displayed on a dashboard are what Tabet describes as particular strengths of eye2. He claims that the dollar turnover of his company is already in “the high six digits,” and puts seven digits to the value of company, though he has not commissioned a formal valuation exercise. “Today, we are still self-financed, but we are considering different options [of future investments from VCs or strategic partners], because we want to expand dramatically,” he says.
Neither company is yet at the stage of offering predictive analytics, but they are working to develop the capacity, with some built-in assurances of customer anonymity. According to Khoury, the English-language ViM is a comparatively affordable solution, because it uses data from security cameras that are already installed in stores. “The data we provide our customers is aggregate data, not specific data,” he says, so ViM could not be employed to identify individual customers and build personal behavior profiles. “We don’t show faces, we don’t show anything [individual]. It is basically a picture of an empty shop with a heat map on top of it. People are already filmed for security purposes and there is no invasion of privacy here,” he argues.
Despite of this limitation by design, ViM rose in less than two years to an estimated corporate worth in the millions of dollars. Khoury says ViM recently did a valuation exercise with a person in the retail industry who was interested in buying a stake or even all of the startup. “Value was estimated to be anywhere between $2 million and $4 million,” he says.
Both startup examples imply that Lebanon-based entrepreneurs can use the future of big data to aim for considerable business success, even if they begin their journeys to big data and predictive analytics while still working on the requisite capabilities. While neither is currently involved in formal valuation exercises with potential investors, both see themselves as million-dollar enterprises in year three after launch, and both are aggressively optimistic about their future. But beyond this sample, what will be the role of big data for local companies?
Some critical notes on big data
The term “big data” has been in use for 20 years and has been defined by three sets of challenges: One challenge is the technical side, where information is accumulating faster. It is today faster than it ever was before, faster and more varied than predicted even during the past decade of fast data growth. In the future, the accumulation of data in terms of volume, velocity, and variety will be greater and faster still.
The second challenge is what to do with the growing resource of raw information: How to analyze it correctly and put the insights to use, and how to protect people from exploitation of their data that is able to overpower them through manipulation or destruction of what used to be their privacy.
The third challenge is the gap between data reality and human assumption, the propensity of humans to think that quantity affects quality. This is what has been called a mythology problem with big data.
New information is supplied much faster than it can be consumed. That is true for unidirectional outbound media (such as this magazine), for bidirectional communication (all verbal and nonverbal chats on mutual media such as communication channels and social networks), and essentially also unidirectional inbound information channels where data from cameras equipped with facial recognition software and computerized communication-monitoring tools amass data volumes that are preserved and remain indefinitely accessible.
In the distributed non-collective data processing sphere of the past few millennia, also known as human memory, the oversupply of such data was a given, and one can blindly assert that it was growing permanently. But it was not a problem because the data from human eyes and ears went no farther than the individual brain, and was limited by a mix of individual data-handling limits (a.k.a. cognitive capacity) and forgetfulness—a blessing, after all.
To tackle the mythological assumption that big data is going to change everything beyond imagination, an ancient paradox is helpful. The sentence, “all Cretans lie,” has been used to confuse unwary students of logic for at least a century. But it is much older. Antiquity attributed the phrase “Cretans are always liars, evil brutes, lazy gluttons” to a pre-Socratic thinker from Crete by name of Epimedes. This dude lived—exact timing is as unknown as the real number of expatriate Lebanese shoppers globally—some five, six or seven centuries before Christ. (The before-Christ part is sure, because the semantic paradox of a Cretan saying that all Cretans lie was quoted by St. Paul in a moralizing tweet to a follower residing on the very island).
Apparently, disinformation is not a new idea at all. Even in the book of psalms, one statement says, “All men lie.” So what is new about big data when the best thing that an economist-cum-data analyst can come up with after several years on the job at Google, sifting through the world’s largest available data stacks, is a 2017 book titled, “Everybody Lies”?
This is not a book review, so issues such as the leading body-perception concerns, unexpected racial stereotypes, origins of media biases, the success of advertising, and questions asked about internet porn by male and female Google users will, unlike in the book by Seth Stephens-Davidowitz, not be addressed.
There is much reason to believe that future influencers will be data-savvier than their contemporaries, and Stephens-Davidowitz concludes that “the future of data analysis is bright,” asserting his belief that every future influencer—every coming Kinsey, Foucault, Freud, Marx, and Salk—is likely to be a data analyst. But so were intellectual influencers since antiquity, from Archimedes and Plato to the original Sigmund Freud and Karl Marx, and certainly also the pre-Google economic shakers and thinkers, from Astor and Vanderbilt to Carnegie, the Hearst and the Koch families, and from Adam Smith to Keynes and Greenspan.
Early in his book, Stephens-Davidowitz concedes that a major reason for the value of Google searches is not so much the large quantity of mined data, but the honesty of people who undertake them, thinking they search unobserved. But will humans adapt their behavior in an age when big data analytics is known to everybody?
One can only wonder when the big data analytics providers for retailers will spot the first data anomalies caused by GPS-tracked legions of pretend-customers, dispatched by interested parties to roam supermarket aisles with the explicit mission of influencing the data that are being collected.
Big data and AI together will improve aspects of economies. But what one should be prepared for, especially anyone presently under 70, is that with the near-term confluence of advanced surveillance technology, and a new intensity of data analytics, the tiny niche of individual human freedom will be even harder to claim than ever.
