For all the recent publication and discussions about the impact that “big data” will have on industries far and wide, the factor which most directly affects the success of these multi-million dollar efforts is decidedly non-technical. Politics is the seminal issue which needs to be handled properly if the promise of “all things data” is ever to be realized.
You heard me right; politics. Not selecting the right Hadoop distribution, not having the perfect configuration of a hybrid public/private cloud infrastructure and not even strong executive sponsorship can help if the elements of corporate politics and the power which they help keep in balance are not openly discussed, accounted for and handled properly.
If you accept the premise that the phenomenon of big data is currently transforming industry; that data is the new oil which will power the next wave of business innovation and success: then one must take a serious look at what the industry is not talking about. Let us not focus on what topics are highly trending; my 9-year-old can spend 5 minutes on Google and find many articles covering use cases of Hadoop and various other No-SQL data stores which are pushing the line of what we thought was possible just a couple of years ago. While these technologies definitely point to a bright future of technical possibilities, they miss the large elephant in the room; corporations are still run by people and people are inherently opposed to relinquishing power.
Data’s Inherent Power Struggle
In his 1997 whitepaper titled “The Politics of Data Warehousing,” Marc Demarest wrote, “Data warehousing projects are frequently side-tracked or derailed completely by non-technical factors, in particular the political treaty lines within the firm, and the politicized nature of data itself.” 
Discussions of “treaty lines” seem more at home in a high school history course than a paper about the great frontier of big data. Yet, anyone who has spent time developing an enterprise data strategy or implementing a large EDW or other analytics platform will tell you the greatest amount of energy was spent dealing with issues of politics and corporate power. From funding, to defining KPIs, to setting project priorities; these are all areas in which political savvy is required, if an effort is going to be successful. Ever tried obtaining financial data for an initiative not supported by finance? Ever see what should be a straight-forward discussion on project status devolve into an argument between finance and marketing about staffing and budgets? Data crosses all departments and data related projects are likely to expose many internal issues.
A Traditionally Taboo Topic
Compounding this issue is the fact that politics, and by extension power, is still one of the least comfortable topics for people to discuss. Rosabeth Moss Kanter stated in 1979 in her famous Harvard Business Review article that “Power is America’s last dirty word. It is easier to talk about money – and much easier to talk about sex – than it is to talk about power. People who have, deny it; people who want it, do not want to appear to hunger for it; and people who engage in its machinations do so secretly.” 
Data also shines a bright light across the entire organization highlighting issues that may have been going on for years. Never underestimate the power of this status quo. A typical data warehousing project whose goal is to create the elusive “single version of the truth” is by definition going to be telling others that their “current version of the truth” is no longer correct, and that their success metrics will now be calculated and evaluated outside their protective oversight.
“Data is much more of an art form, orchestrating people and their politics.”
A CMO who has always been responsible for calculating their return on promotional investment now is being told what their performance metrics will be. The head of sales who has been calculating the incentive compensation numbers for their entire sales force now is being told what the new performance metrics are and how they will be calculated. In order to be successful data practitioners, we should first obtain an understanding of the existing political landscape, develop a sense of partnership with these key stakeholders, and then help these executives see the benefits of moving forward together into this great new world of data-driven business. Anything short of full inclusion and trust of key stakeholders will result in a political power struggle which will slow down progress and can even stop the best technical solutions in their tracks.
While it is tempting to view data science in all its forms as a quantitative and engineering focused field, my 20 years in the trenches have taught me that data is very much a social science. Data is much more of an art form, orchestrating people and their politics, than an engineering problem to be solved. Learn how to harness the inherent politics of corporate America and see your data initiatives flourish.Footnotes
- Marc Demarest, “The Politics of Data Warehousing”, June 1997
- Rosabeth Moss Kanter, “Power Failure in Management Circuits”, Harvard Business Review, July 1979