15 December 2019

Can We Get Social Media to Work for Society?


As debates rage over the need to regulate Facebook and other social media giants, a question that keeps coming up is whether these platforms can be regulated without stifling innovation. In this opinion piece, Ravi Bapna argues that the answer may lie — among other factors — in a combination of business model innovation and creative regulation. Bapna is a professor of business analytics and information systems at the University of Minnesota’s Carlson School of Management.

If you feel overwhelmed by the deafening din from the punditry — that includes just about anybody from former founders to lawmakers to presidential candidates — about how toxic Facebook is, you are not alone. The litany of issues, questions and its responses when lined up are as dizzying as they are telling of the times in which we live.

Should Facebook — and other tech platforms — be regulated? CEO Mark Zuckerberg says it should be. Should the regulator be Congress, the Federal Communications Commission or both? Should Facebook, WhatsApp and Instagram be broken up? Is Facebook a platform or a publisher, and thereby subject to, say, libel laws, or free equal airtime laws for political candidates?


Facebook has conveniently argued on both sides of this issue. If a politician gets airtime on Facebook’s wall — as governed by its engagement maximizing algorithm — how is that different than, say, Fox giving Bernie Sanders free airtime under the equal opportunity statute because it gave Trump the same? How did the company survive Cambridge Analytica, Russian trolls, Iranian trolls, Alex Jones, fake news, the slaughter of Rohingya muslims? Is the Zuckerberg-Sandberg duo going to end up on the right side of history? I won’t bet on that one. Will it follow Jack Dorsey and Twitter to the high road — shall we call it a bylane given how small a share of total advertising it is — and ban political ads? Wait, Washington state, with some of the toughest campaign finance laws in the country, already prohibits political ads on social media, but why did Facebook not enforce this?

Publishing Policies Versus Profits

Yet, contrast this debate over Facebook’s publishing policies with the company’s third quarter 2019 earnings reports. What do we learn about ourselves as a society when we see that revenues are up 29% year-on-year and net income is up 19% to $6.1 billion? I believe that reflecting on this question can lead us to a productive dialog on how we can get social media, a powerful modern technology that powered the Arab Spring close to a decade ago, and still today is helping address some of worst repression and income inequality issues in Latin America, to work for society at large.

Univison’s Jorge Ramos rightly reminds us that official communications (from increasingly authoritarian Latam regimes) are balanced by millions of videos, photographs and texts on Twitter, Instagram and Facebook. In doing so, we can take a longer view and distance ourselves from the specifics of the platforms, CEOs and idiosyncrasies of the day. These too shall pass, as the old saying goes. And, as I have mentioned in this very forum, much good is still likely to emerge from a shrunk world that has billions of people socially connected to one another on digital platforms.

My view is that it will take a village for this to happen. We cannot leave it to the platforms to self-govern. Nor, based on what we have seen in Congress so far, can we expect miracles from lawmakers, who seem to be perennially behind the curve when it comes to digital technologies and their societal consequences. In equal part, I believe, it will take agency on the part of end users and corporations to be more thoughtful and rigorous about their value calculus from social media platforms. My conjecture is that both end users and corporations are over-estimating the value these platforms provide, but the pathways and mechanisms that lead to this bias are different. On top of this, we will need technical and business model innovation coupled with creative regulation that encourages interoperability and competition in a context that naturally favors network effects based winner-take-all economics.

“Much good is still likely to emerge from a shrunk world that has billions of people socially connected to one another on digital platforms.”

From the users’ perspective, it’s a tautological fact – nicely captured in an old Hindi saying that roughly goes like ‘if there is no flute, then there is no symphony’ – that the entire business model of purely ad supported digital platforms such as Facebook disappears if users decide that the benefits are lower than the costs of using such platforms. To answer the question why is it that people can’t resist checking how many likes they are getting on their recent Facebook post, or signaling how hip and happening their social lives are on Instagram, we need to perhaps ask, what real world social vacuum are these platforms swarming into? If kids are addicted to devices, is this a result of the technology or is it cop-out parenting, which may be a result of weakened social fabric, or the ever-elusive work-life balance, that the adults face? Or does the problem lie in the lack of engagement at school, which in turn could be driven by shockingly underpaid teachers? Users will benefit from examining the true costs of their online interaction. It will give a more accurate baseline against which they judge the true value of social media to themselves.

The same logic applies to organizations pumping billions of dollars of advertising dollars into platforms like Google and Facebook. These organizations need to get more rigorous about estimating their ROI from these digital investments. A recent summary of a decade of rigorous, gold-standard RCT based, research into the effectiveness of digital advertising largely concludes that it’s more like the next dot-com bubble. This is a quarter-trillion-dollar market that gives the illusion of measurement but is still governed by irrationality of empire-seeking marketers who measure their self-esteem as a function of the size of their organization’s marketing budget.

When economist Steve Tadelis and colleagues designed a counterfactual experiment and shut down branded keywords advertising on Google, the users still found eBay via the organic links. Ebay was literally burning $20 million in targeting ads using the word ‘eBay.’ For every dollar eBay spent on this sort of advertising, they lost 63 cents. Similar studies on Facebook have found that traditional correlational measures (the type most often used by practicing managers) over-estimated the true causal efficacy of a retailer’s ad by a factor of 10. Scale this out to the $275 billion digital ads industry and you can get a sense of the societal problem. On top of this is the problem of fake ads, especially in the world of display advertising. NYU scholar Anindya Ghose recently quantified the scale of this problem saying one in five ads one sees on a smartphone are fraudulent. My point is that if marketing departments in tech companies (eBay for example, prior to Tadelis’ study) are grossly incapable of doing causal analytics, of exercising counter-factual thinking, and thereby are, day in and day out, over-estimating the true effectiveness of their digital ad dollars, think about the scale of this problem across the global economy.

“Imagine a world where users get savvy about their own utility from platforms such as Facebook, and advertisers scale down their Facebook ad dollar investments by, say, a factor of 10.”

Next-gen Business Model Innovation

Now imagine a world where users get savvy about their own utility from platforms such as Facebook, and advertisers scale down their Facebook ad dollar investments by, say, a factor of 10. This will be when genuine next-gen business model and technical innovation will accelerate for social media. On the technical side, newer capabilities in smart contracts that go hand in hand with scalable and real-time blockchain based verification will create avenues for individuals to get compensated for the usage of their data and for the veracity of digital ads. On the business model side, as I have argued before in this forum, there is opportunity to move away from purely ad supported social media to subscription based or Freemium based privacy protecting and ad-free social networking.

Avinash Collis, Erik Brynjolfsson and Felix Eggers have estimated that Facebook is worth about $40 to $50 a month for U.S. consumers. This should be encouraging news for any startup, or say a tech company in Redmond, WA, looking to create the Netflix of social media. It could also be a welcome challenge for someone who has a genuine focus on using artificial intelligence, machine learning, social and digital for improving the human condition.

No comments: