Our Thinking - What is a data-driven culture? (Part 1)

What is a data-driven culture? (Part 1)

Many firms are wondering how to become more data-driven in their approach to solving problems. More often than not, these firms first focus on and invest in technology solutions to give their data meaning and provide deep insights. If you are a leader and you are looking to become more competitive in your industry through becoming more data-driven, do you ever wonder if your competitors are thinking and doing the same thing? How much are they spending on technology in order to crunch more data and deliver deeper insights than you? What is a competitive advantage if all your competitors are fighting with the same technological weapons? You could view Business Intelligence (BI) as just another thing you must now do as a “ticket to the game” in order to remain competitive. However, if you want something close to a sustainable competitive advantage perhaps you should look to the unique culture of your firm and try to influence it to become more data-driven in its approach to solving problems. After all, a critical component of a successful BI or analytics project is the effectiveness of the organization that consumes the data. The Return in the ROI formula for BI systems are almost solely based on the quality of the organization using the new technology and the culture that underpins it. The technology is the investment, the return is how you capitalize on the investment into making data more relevant and consumable.


Why is culture an important competitive advantage? Because a high performance culture is extremely hard to replicate. We’ve all read Good to Great by Jim Collins about how organizations who enjoyed sustainable out-performance over time had something unique in their DNA that their competitors did not. So what is a culture and how do you influence it? First of all, I hate the word culture. Sure, we all kind of  know what someone means when they talk about their organization’s culture but it just always seems so nebulous and hard to decompose into its true meaning and features. I believe culture is essentially the organization’s three-legged-stool of decision rights, performance appraisals, and rewards system. In essence — accountability.

Three Legged Stool

This was one of the foundational concepts pounded into my head for two years during my MBA studies at the University of Rochester Simon School of Business. It is also something I spent a lot of time observing during my years at Xerox Corporation. The organizations that performed the best were those with a clear pattern of accountability. In a highly competitive and political environment such as Xerox there were few things that mattered more in the corporate “Art of War” than maintaining a cultural advantage. All else being equal, no matter how smart your opponent or how much resources he has, if your organizational machine outperforms him — you win.

You don’t influence an organization’s culture by changing a mission statement. You drive behavior and accountability to be aligned with your mission and vision. If you have a vision of transforming your organization to be more data-driven think first about what behaviors you wish to influence and how to influence them. Ponder how you will allocate decision rights in your organization (who will be making data-driven decisions), then how will you evaluate people on the quality of the decisions they made, and lastly how will they be compensated. If there are no rewards for excellent performance or consequences for poor performance then you have a serious accountability issue on your hands…and the culture will suffer.

Decision Rights

When I think about decision rights I think first about centralization vs decentralization. This may seem like a simple task but there are religious debates and ivory tower scholarly articles raging about how to get those in the organization with the most tacit knowledge and experience the rights to make decisions. Basically, the answer of centralize or decentralize is “it depends”.  Fundamentally, to make a decision you need information.

Centralization means that you will incur costs in collecting information from those who have it and transmitting that information to those who have been chosen to centrally make decisions for the organization. You may be able to justify these costs if the decisions that need to be made must follow strict guidelines or require specialized knowledge and qualifications.

Decentralization means that you have decided that the costs of information transfer and loss would be too great in a centralized model, and the risk of allowing your individual employees and work teams to try and fail is sufficiently low.

A mortgage company for example may have relationship managers that get to know their clients and try to meet there unique needs, but at the end of the day the underwriters control the risks to the firm so that individual relationships and preferences don’t end up becoming a risk to the entire organization. In contrast, Amazon.com has developed the “two pizza” model wherein teams, no larger than can consume two pizzas, are empowered to solve problems, and to try and fail quickly. If someone at Amazon has a data-driven solution to a customer problem they can prototype a solution, beta test it, and decide whether to roll it out more broadly on the site with very little cost and consequence. (see http://theagileexecutive.com/tag/two-pizza-team/).

It all boils down to where you believe the valuable information exists in the organization. Is it at the fringes — the first line? Or is it in pockets of specialized people and teams?  Could it be that valuable information exists everywhere in your organization in those that have the most tacit knowledge on their subject area. Once you decide this and weigh risks and costs you can determine who should be able to make decisions based on the information they posses or know best how to interpret.

In a future post I will continue to expand on this and discuss methods for influencing behavior through accountability systems (appraisals and rewards). In addition, we will add more definition around what it means to be “data-driven”.


Submit a Comment

Your email address will not be published. Required fields are marked *