Data analytics and social networks increasingly drive the digital transformation of the firm, says Eric Clemons, a Wharton professor of operations, information and decisions, in this opinion piece.
For corporations, the keys to successful change lie in the digital transformation of strategy and organizational structure, he also notes. Yet, while the general direction of the digital transformation wave is increasingly clear, successful implementation “throughout the organization is far more complex.” Below he offers ideas about how to approach the challenge.
We are immersed in a sea of information, provided by cheap and powerful information technology. For the first time in human history, most of us no longer think of information as an asset or a liability, or as a source of power or of weakness. Instead we think of iPads and Droids, Google and Apple. Mostly, we no longer think about information at all, any more than fish think about the water that surrounds them. But information tides, information currents and information tsunami affect us in our information oceans as profoundly as tides, currents and tsunami affect the fish in their oceans. We really have begun the digital transformation of everything.
At regular intervals, technology makes enough progress that we in the punditry believe that technology now must now transform management, business and politics. High-speed data communications have indeed altered the business of securities trading beyond recognition, and have given individual traders both financial rewards and the ability to bring the global economy to its knees from time to time. What about the impact of database management? What about the impact of model-based management? What about wearable computing devices? Individually, each has had at least some impact. But collectively they are now producing the ultimate in corporate transformations, the Digital Transformation of the Firm.
Why is Digital Transformation Interesting Now?
There are two trends, and they are pulling every firm in two directions at the same time.
- Big data analytics: It looks at everything. It looks at information the firm already has, so it is available almost immediately, and it is not subject to a number of the limitations associated with survey data. It is often based on machine learning — using the data to discover patterns that no expert might have predicted — rather than on traditional hypothesis testing, so it shifts moral authority away from traditional experts and towards the technicians most able to crawl through and analyze truly enormous data sets. And these data sets truly are too large to share, so they create local fiefdoms around the sites where terabytes, petabytes, exabytes and zettabytes of data are stored.
- Online social networks: They potentially connect everyone, but we know from social sciences research that they do not. What they do do, and they do it well, is strengthen existing connections among individuals and groups, and allow those groups to remain connected over time and across time zones and huge spatial distances. Where big data analytics may create local fiefdoms, online social networks create distributed pockets of autonomous connection, affiliation and even affection. Digital natives place different values on data driven analytics from big data people, but they have a similarly dismissive attitude towards traditional “old guy expertise.” I have had students in my MBA classes text each other and vote on what they believed to be the correct answer to a discussion question in class. I have had an online reporter for the Seattle Post-Intelligencer explain to me that a hive mind may not be deep in knowledge but it is broad enough to be self correcting, and the views of 500 members of her hive were more reliable than any small group of experts.
“With different information flows, organizations almost certainly will adopt different designs. The most basic examples involve centralization or decentralization of decision rights.”
The first trend divides the firm by employees’ locations, while the second divides the firm by their affiliations. Surprisingly, neither of the two digital technologies at present actually integrates the firm. One values data-driven analytics, and the other values consensus and shared experience. Surprisingly, both devalue traditional experience-based expertise. Both are real, and both are becoming more important. And both will affect the strategy and the structure available to the firm.
Both are driving a new digital transformation.
What is Digital Transformation of the Firm?
There are two distinct components.
The first component is the Digital Transformation of business strategy. There are new gateway-centered businesses, like Google, that can advance or destroy almost any company simply by facilitating their access to customers or by cutting them off. The power of these gateway-centered businesses cannot be overstated, as evidenced in their ever-expanding profits and their ever-expanding legal difficulties.
But even the most mundane businesses, like consumer credit cards and fast-moving consumer products, have been transformed as well. For example, in credit cards decades ago all issuers charged all customers the same annual fee, and all issuers charged all customers the same interest rate on their finance charges. Today the best card issuers offer thousands, even tens of thousands of different combinations of annual fees, interest rates, cash back rewards and other incentives. And the best issuers use sophisticated models to determine which existing customers and which new card applicants should receive which offers, in order to maximize the risk-reward profiles of the issuer’s entire portfolios.
A couple of decades ago we all drank products from Budweiser, Miller, or Coors, and it looked like we were going to have no other choices. Beer would be adequate but mediocre, and success would be determined by a beer company’s advertising staff, not by its brewers. Today, we have tens of thousands of beer choices in America, and a company’s success is largely determined by the quality of the beer itself; social networks, online reviewing sites, user-generated content, word of mouth, and word of mouse have replaced advertising as keys to the success of new product offerings. This is true not only in beer and soft drinks, but in a range of new resonance marketing products like snacks, blue jeans, coffees and teas, artisanal breads, and sea salts. Gateway businesses, differential pricing, and resonance marketing are only three of most prominent changes resulting from the digital transformation of business strategy.
The second component of the Digital Transformation of the Firm is the Digital Transformation of the structure of the organization itself. This is best understood through historical examples. When Benjamin Franklin was the American Colonies’ ambassador to France in 1776 he had enormous discretion in determining his actions, which were directed at gaining critical French support against the British in the Revolutionary War. This personal discretion was inescapable, despite or even because of the critical importance of his mission. Communications between Franklin in Paris, and the American capital in Philadelphia took weeks to accomplish. In contrast, American ambassadors today are subject to much tighter control from Washington because secure international communications are almost instantaneous.
Snipers since the introduction of the rifle have had enormous autonomy, because only the shooter on location had sufficient information to select a target and execute a mission. In contrast, today’s drone attacks are conducted from half a world away, by a flight control with far more information than could ever have been made available to a sniper even a decade ago.
“Apparently, executives responsible for VW’s diesel products, executives responsible for VW’s North American sales, or both decided that their best career moves would involve faking their test results….”
The design of the firm, then, can be seen in terms of three components:
- Decision rights: Who gets to make each critical decision facing the organization? This should be dictated in large measure by who has the information necessary to rapidly decide what should be done, and by who has the ability to rapidly implement a decision once it has been made.
- Property rights and incentive systems: How do we measure performance? How do we reward performance? Historically, the owners of the firm received the benefits from good decisions, and endured the costs and the risks of bad decisions. The most effective way of ensuring that their agents — their employees and their managers — made the right decisions was to let their agents “own” some of the results of making decisions. This could lead to agents taking actions that were best for themselves or their own divisions, but less than optimal for the organization as a whole. This in turn led to the inclusion of more broadly based incentives and rewards systems, like annual profit sharing plans.
- Information flows: Who knows what? Who knows it when? How quickly do they know? How quickly does information flow up and down and laterally across the firm?
With different information flows, organizations almost certainly will adopt different designs. The most basic examples involve centralization or decentralization of decision rights. In the case of critical decisions, like the role of foreign diplomats and ambassadors, more rights have been claimed for the central organization as information speeds, accuracy and security have all increased. When the cost of a short delay in deciding and acting is lower than the cost of mistakes, as is the case with drone attacks on targets embedded within civilian populations, increased central control is generally optimal. In other cases, like dogfights among pilots in supersonic fighter aircraft, local autonomy of the pilot is the only way currently available to increase pilots’ chances of survival.
In theory, the progression in organization designs should be clear. When information flows were limited and operations of the firm were relatively simple, international businesses were global. Products were designed centrally, and often were manufactured centrally and then exported for sale through captive organizations. General Motors designed and manufactured cars in North America, and then exported them for sale around the world. As local design to meet local preferences became more important, firms became multi-domestic. They had local operations, designing and manufacturing for local populations, and marketing and selling to local populations. We now expect to see more integrated multinational corporations, designing a family of different but related products for local markets, manufacturing wherever costs are lowest and quality it highest, and selling through local organizations, but all subject to close oversight from the firm’s global headquarters.
In practice, we’re all still learning how to do this. It’s most striking when a great firm gets the design really wrong, as we saw with BP’s Deepwater Horizon Gulf of Mexico disastrous oil spill and more recently with Volkswagen’s diesel emissions control scandal.
In the case of Deepwater Horizon, it appears that BP’s incentive structure was deeply flawed. BP allowed the safety decisions for the rig to be made by local managers, who presumably had access to better, more timely and more accurate information. These managers felt that the penalties for keeping the rig off-line while normal start-up checks were performed were too high for them to accept. The rewards for bringing the rig back online sooner were too high to resist. And the penalties for a blow-out, while catastrophic for the company and for the residents of the Gulf, were not much worse for the managers themselves than those associated with leaving the rig off-line for normal safety checks. The results of this poor organizational design are now clear.
“The general directions of Digital Transformation are clear to all of us. The successful implementation of Digital Transformation throughout the organization is far more complex.”
In the case of the Volkswagen diesel emissions scandal, senior management once again created a seriously flawed organizational design. The diesel product division was given clear targets, including rapid growth in the United States. Since VW was counting on their diesel models to achieve the necessary fuel economy targets for their U.S. fleet while meeting U.S. EPA air quality standards, this required that the diesel fill a niche comparable to that of the hybrid electric models of Japanese sellers such as Toyota, Nissan and Honda. While it is not yet clear that a diesel-powered car cannot fill this role, it is now clear VW engineers have been unable to design such a diesel to-date. Apparently, executives responsible for VW’s diesel products, executives responsible for VW’s North American sales, or both decided that their best career moves would involve faking their test results, fraudulently passing their US test results, and selling cars that were far worse polluters than they appeared to be.
Once again, the company’s incentives were clearly wrong. As importantly, information flows were clearly defective. How could VW’s most senior executives not wonder how VW, and VW alone, had mastered a super-clean diesel, with pollution levels so much lower than they had been just a few years earlier and so much lower than any of their competitors? How could they not have enough information to verify how the engine had been redesigned to achieve these results? How could cheating on such a massive scale not have been detected by these executives?
Clearly, the social network of those involved with VW’s Diesel operations — those associated with meeting sales quotas, assessing the capabilities of engine design and the prospects for improvements, and those associated with passing environmental testing requirements in the U.S., Europe — and indeed around the world — shared beliefs about protecting themselves and their group. Clearly, they believed in the mission of their group above the mission of the corporation. And just as clearly, VW’s senior management, including members of the board, were not part of this social network and were not aware of the deliberations or of the decisions of this social network.
It’s clear that Digital Transformation must be embraced by the entire organization. A move from a mass-market orientation at Anheuser Busch to resonance marketing at its Goose Island and Six Point operations cannot be made solely by digitally savvy employees in the field. A move from rigid central control to autonomous operation of drilling platforms clearly is disastrous without the right incentives and policies in place. A move from rigid central control to autonomous profit maximization by region or by product group cannot be made without careful monitoring of incentives and of detailed and accurate information flows.
The general directions of Digital Transformation are clear to all of us. The successful implementation of Digital Transformation throughout the organization is far more complex. As more and more of the New Digital Natives — the new Petites Digerati — assume positions of authority within their organizations, it becomes increasingly important to understand their thought patterns, their interpersonal style and their very different patterns of loyalty. This is one reason why my team now includes an anthropologist, a psychotherapist, an economic historian and an international negotiator, where it used to include specialists in telecommunications, algorithm design and database management.