Mindset within ‘smart data’ & human edge
Digitization – with innovation and scalability as the pillars of competitiveness – is based on the rapid development of IT, which made ‘smart data’ possible and enables scaling with little capital in nearly no time and, which led to massively increasing global flows of data, finance, talent and trade.
Digitization is shaping new frontiers of value for customers and thus is blurring boundaries between industries and is giving rise new and more global competition – potentially culminating in pure-play disrupters or ecosystem shapers (cf. Dobbs et al. (2014), Dörner & Edelman (2015), Catlin et al. (2015)). Transparency and marginal costs partially tending towards zero lead to pressure on prices and margins and finally on commoditization and consolidation, and life cycles of companies are shortening (cf. Hirt & Willmott (2014), cf. Rifkin (2014)).
- Consumer trends – including gadgetry – range from immersive experience (augmented and virtual reality), smart assistance (artificial intelligence) and gravitational (personalized) content to instant access (boundless information and instant communication) and platform economies (sharing and using instead of possessing) (TRENDONE GmbH (2016)).
- Conventional ‘pipeline’ businesses are complemented or substituted by platform businesses, which “bring together producers and consumers in high-value exchanges” (Van Alstyne et al. (2016)). Such platforms challenge fundamentals like scarce resources with clearly defined ownership rights. In order to create user traffic, they offer open architecture that grants access to the platform resources, and open governance that allows players other than the owners to shape the rules of trade and a ‘fair’ reward sharing system (Van Alstyne et al. (2016)). The network effects, i.e., the demand-side economies of scale allow a platform to almost always win when entering a pipeline firm’s market and in some markets to even create a winner-take-most situation (Van Alstyne et al. (2016), Henke et al. (2016)).
In this environment of more data but more discontinuity, less certainty and more volatility, the conventional corporate governance model has to be questioned regarding its adequacy to foster the digital transformation by focussing sufficiently on the viability of the business, e.g. with respect to advanced analytics, digitizing or automating supply chains and other information-intensive processes or robotic process automation (cf. Dahlström et al. (2017)). And, whereas a tight control of internal resources dominates in conventional ‘pipeline’ businesses, in platform businesses leadership needs to nurture external ecosystems (Van Alystne et al. (2016)). Thus both, management and board of directors, have to adopt their mind-set and to adjust to changing leadership skills, particularly with respect to “rewiring the mechanisms for making decisions and getting things done” (Dahlström et al. (2017)): Computational analytics for decision-making versus creative judgement as means of differentiation and social network diagnostics complementing inspirational leadership in execution determine the areas of tension.
‘Smart data’ promote computational analytics as a basis of – partially automated – decision-making at higher speed. There is no doubt that combining vast amounts of data and increasingly sophisticated algorithms objectifies decision making by refuting biases that cloud judgment, by enabling more accurate predictions, and by opening up new pathways for performance optimization (Lovallo & Sibony (2006), Rosenzweig (2014)). Moreover rigorous (real-time) data monitoring allows to quickly refine or to jettison initiatives (Catlin et al. (2015)). Corresponding models increasingly cover all areas of (business) life: they range from real-time information about customer behaviour, to monitoring credit card use and to approving loans, or from weather simulations targeting improved crop yields to forecasting the quality of wine vintages (Rosenzweig (2014)), or even from the analysis of opposing sport teams. The growing value for management is apparent and acknowledged. On top of this, social (media) tools, like online conferencing, social networks, wikis, podcasts and blogs, are transforming businesses by speeding up the access to knowledge, reducing communication costs and facilitating greater participants’ involvement. They could further induce data-driven decisions or even ‘democratize’ decision–making (Bughin et al. (2015a)). Thus, computational analytics and social media will substitute management intuition in many respects (Dobbs et al. (2014)). But it will require a more and more data-analytics oriented strategy to ensure high quality, i.e. validated and differentiated inputs (for artificial intelligence and machine learning) as well as prioritization of data and their flagging for escalation (Dewhurst & Willmott (2014)). The imperative to ‘let go’ and “devolve decision-making authority up and down the line” becomes more compelling: information, i.e. valuable insights, needs to be democratized (rather than bureaucratized) to enable “the organization to manage itself without bringing decisions upward” (Dewhurst & Willmott (2014)).
However, ‘blind faith’ is one of the pitfalls when using ‘smart data’ and social tools for forward guidance and decision-making. Applying sophisticated and – even more so – ‘smart data’-driven models bears the risk of “a ‘shut up and calculate the numbers’ ethos …, as if technical proficiency with programming could substitute for thinking about what the numbers mean” (Fox (2011)). Plus, there is the danger of the general opinion manipulating one’s mind via social media due to the herd instinct of the human, which could lead to ‘lemmings’ decisions’ and ‘jumping on the bandwagon’.
At the same time, differentiation needs creative judgement. As logic is indivisible (i.e. same information should lead to same conclusion), ‘egalitarianism’ (of systems) due to infinite transparency and the resulting intensified form of commoditization represents an increasing challenge in a world dominated by ‘smart data’, available at continuously decreasing costs and shortening time lags. To put it in an extreme way: every solution, even for problems under uncertainty, will be programmable as an algorithm and will be even more readily and unrestrictedly accessible on the market. As a consequence, as all have the same access to speed, winning will only happen in turns, demanding ‘ingenious’, surprising moves in order to remain at the forefront of competition, alongside with the agility to change or reinvent quickly and the capacity for self-renewal. Amazon may be an excellent example for expansion by always looking for new business ideas (ranging from developing hardware products to streaming contents) and combining pipelines with platforms. The necessary creative judgement requires tolerating ambiguity (and “delaying decisive action until clarity emerges”), encouraging ‘exceptions’ which “may pave the way for innovation”, and establishing “a set of small, often improvisatory, experiments to get a better handle on the implications of emerging insights and decision rules” (Dewhurst & Willmott (2014)). Great surprise creates change beyond the fact itself, be it either as advance or lesson learnt. But with increasing impact of creative judgement, embedding debiasing techniques in the formal decision-making process gains importance. Scenario planning, vanishing-options test, ‘devil’s advocate’, ‘war game’ or ‘pre-mortem’ represent examples of such techniques (Meissner et al. (2015)).
Social networks do not only change communication and interaction, i.e. connectivity in general and staff leadership in particular, but also allow for social network diagnostics to create personal profiles with formerly non-measurable information on personal characteristics and capabilities, processes and behavioural practices (cf. Brower-Rabinowitsch & Hergert (2015), Bughin et al. (2015a)). This could lead to key changes not only in recruiting, but also in organizational structure and management processes; as mentioned, democratizing decision–making or promoting the use of internal markets and voting mechanisms to allocate resources, or flattening hierarchies/making them disappear (Bughin et al. (2015a)). Thus learning organizations have to use people analytics to improve collaboration, to build teams and to create momentum for action and change.
The human edge still excels artificial intelligence, the more emotional intelligence (persuasion and empathy), i.e. inspirational leadership and in particular role modelling (and mentoring) is needed. This refers to the contextualization and enculturation of ‘smart data’ analytics, both with respect to decision-making and social diagnostics, but even more in general when it comes to execution. Rosenzweig (2014) argues for the need to be cognizant of the limits of decision models via the “distinction between outcomes leaders can influence and those they cannot”, i.e. the difference between making things happen versus predicting what will happen beyond the own sphere of influence. In this context emotional intelligence often also serves to reduce complexity. Impact results from actions, with leadership in execution being a core skill to limit implementation gaps, i.e. ‘governance arbitrage’ losses.
Execution, in connection with automation, will benefit from decision models the more those will become decision making models, i.e. shifting from indirect to direct influence, like autopilots or automatic stock trading. Likewise, for change management the further development of social (media) tools will be essential. Against this background, management and board of directors have to develop an open mind-set for the productivity evolving from ‘smart data’ by being technology savvy, while making conscious use of their human edge in order to innovate and differentiate, and they have to adapt their roles and agendas accordingly (cf. Dewhurst & Willmott (2014)). “The emergent nature of digital forces means … a relentless leadership experience and a rare opportunity to reposition companies for a new era of competition and growth” (Hirt & Willmott (2014)).
Frankfurt, 27 March 2017
 Not only that ‘the fast ones eat the slow ones’, but also growth is a key driver for valuation and as such also can materialize as acquisition currency.
 For example, since 2005 a team of researchers from the German Sport University in Cologne supports the German national soccer team with dossiers on their opponents (Kaufmann (2014)).
 As an analogy to the theoretical end state of information egalitarianism, physics describes the maximal entropy, i.e. the so-called ‘heat death’ or ultimate fate of the universe (cf., for example, Baestlein (1991)).
 In competition the credo must be staying unpredictable (not to be confused with unreliable) in an internally measurable and calculable way. Like leading soccer teams shift away from the one system how to play the game to a multifaceted, flexible and opponent targeted approach with studied moves and behavioural automatisms as enablers (cf. Eichler (2014)). A musician could describe this as spontaneous improvisation based on elaborated compositions.
 Meissner et al. (2015) propose a decision-screening matrix – to be applied from the outside – with consideration of downside risk (vs. risk of overconfidence) and of different points of view (vs. risk of confirmation bias) as dimensions.
 Social network diagnostics potentially comprise human-resource data analytics, like social network fingerprints, as well as social media technologies, including voice and face analytics or eye tracking.
 Dewhurst & Willmott (2014) emphasize the growing importance of the softer side of management vis-à-vis ‘thinking machines’ because it’s the least replicable by algorithms and artificial intelligence. Maskin considers that “humans are instinctively moral beings and … [does not] see machines as ever entirely replacing those instincts. Computers are powerful complements to moral reasoning, not substitutes for it” (Maskin & Winter (2016)).
 Grove (1997) calls the divergence between statements and actions “strategic dissonance”.
 Rosenzweig (2014) illustrates his perspective on combining analytics with a performance-oriented mindset by the example of ‘Moneyball’. ‘Moneyball’ describes, “how the Oakland Athletics, a low-budget [baseball] team in a small market, posted several consecutive years of excellent results” through assembling a team based on relying on decision analytics (‘sabermetrics’) how to “optimize runs scored per dollar” (cf. Lewis (2003)). However, once on the field “the reality is different. Players don’t predict performance; they have to achieve it” (Rosenzweig (2014)).