Skeptics or Suckers? How to overcome executive and IT resistance to data-driven decision making

Guest post by Kevin Gray, edited by Koen Pauwels resistchange It took the human race tens of thousands of years to put men on the moon. Modern science is only a few hundred years old, so it should come as no surprise that scientific decision-making does not come easily to human beings. Indeed, business decisions often seem to be made on the basis of gut feel, organizational politics and simple accounting.  Hype sells better than science[1], as demonstrated by the confusion of what e.g. neuroscience reveals about consumer behavior[2]

Managers often have had no more than the compulsory “Statistics 101” as part of their formal education, which may be why simple descriptive statistics such as the number of Facebook likes are sometimes presented as “advanced analytics.”  Elaborate, sexy graphics are often conflated with statistical methods.  Confusion of software, programming and other facets of computer science and IT with analytics is widespread and buyers of “analytic solutions,” swamped by well-packaged claims, may not always be getting what they thought they had bought.

Faced with tight budgets and resources, managers understandable fear the unknown.  At the extremes, there seem to be two camps: Skeptics (“I’ll believe it when I see it”) and Suckers (“I’ll see it when I believe it”). It’s natural to defend the status quo and protect one’s turf but resistance can harden in the face of sales pitches that border on science fiction, and it may take only a few bad decisions by major companies to turn key influentials off to the very idea of analytics-based decision-making. Necessity may be a mother of invention but hype is surely one of its hangmen!

In our experience, Skeptics typically outnumber Suckers in the C-Suites of most companies. However, many senior leaders feel pressure from shareholders and opinion leaders to get on board and at least put on a progressive face. A decision to publicly embrace analytics may be made hastily, on the basis of gut feel without considering costs as well as benefits. Obviously, this would be in opposition to the central idea of analytics-based decision making!

Many decision-makers have a lot of data at their fingertips or easily acquired but they and their teams do not know how to use it effectively. Part of the reason goes back to human genetics but it also reflects corporate culture, which varies more than our genetic makeup. A radical change in the way decisions are made cannot happen overnight, except in very small organizations and even in those rare cases it is easier said than done. It requires more than a few policy revisions – a change in corporate culture is needed before analytics and data-based decision making can be truly in place and up and running.

IT is a key area for change and in some organizations perhaps the focal point. “They never notice us until the server goes down” is a grumble that has been made by more than one IT head. It’s a tough job and many IT departments are overstretched and underfunded, and the last thing they need is Big Data (in some cases, even more Big Data). Sometimes, they lack the necessary IT skills in-house and are reluctant to outsource or are unable to, owing to budget constraints. Furthermore, IT professionals are computer science professionals and not necessarily well versed in research methods and statistics, which are not core IT skills. IT duties themselves comprise a full-time job and getting the data infrastructure functioning for decision makers and their staff in various parts of the organization that have diverse needs is not trivial. The data, tools and internal client needs also are not static.

There are turf builders too, naturally, who welcome the opportunity data and analytics present but many of these more ambitious types do not really understand how to integrate sophisticated analytics into decision-making.   For example, Support Vector Machines (SVM) may or may not be the best technique for a given situation, and another consideration is that this (and other methods) come in many flavors – SVM does not refer to a single procedure that can be mechanically applied. These sorts of distinctions are second nature to formally trained and experienced data analysts but can strike whose who are not as waffling. One SVM is the same as another, as far as they are concerned. So, while turf builders may be enthusiastic about adopting or expanding analytics, they may not have a good grasp of what it is and the skills and resources needed to make it work. Is IT is a barrier or a catalyst for change in data analytics use? Very much depends on the organization. Ideally, professional data analysts should be able to extract the (clean) data they require in order to address a specific business question in a straightforward manner. They need to be able to do this proactively as well, to identify potential threats and opportunities and make the appropriate recommendations to management. However the data ecosystem in which we all live and work is constantly evolving so there should be no surprise if there is resistance from IT.

It will usually not be prudent to build an in-house analytics team primarily around the IT function. Diverse skills are needed but all who are part of the team (including dotted line reports) must first and foremost understand what analytics means and how it can be used to enhance decision-making. This usually amounts to a corporate cultural change. CEOs should not just dump expansive new responsibilities on others and then walk away – senior management must be involved hands on from start to finish for this sort of initiative to pay off. Though there are many kinds of technical expertise needed, this is not just a matter for techies to sort out on their own. Ideas are often the easier part of innovation – successful implementation can be much harder – and once the decision is made for a significant investment there may be no turning back. Until the train derails.

What is your experience with executive and IT understanding versus resistance regarding using data analytics? Please let us know!

[1] http://www.greenbookblog.org/2014/04/14/innovation-or-sales-pitch/ [2] http://www.greenbookblog.org/2015/02/24/how-to-separate-neuroscience-from-neurohype/?utm_content=buffere2ca8&utm_medium=social&utm_source=linkedin.com&utm_campaign=buffer

Advertisements

Leave your comment here

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s