18 Dec 2018Blog

One does not simply build a data strategy without considering culture

Technology alone is not a strategy. Even if it was, it is no match to culture.

Peter Drucker emphasized the importance of culture when he famously said “culture eats strategy for breakfast”. Yet, we see a number of “strategic” programs focusing on installing new technology, without considering that technology alone is not a strategy, and even if it was, it is no match to the culture. There is plenty of research indicating that culture and people are the biggest obstacles for success in data & analytic initiatives [1]. Therefore, cultural considerations are necessary if you want to build a successful strategy for becoming data-driven.

There Is a Culture Gap Between “Old” and “New”

Among many definitions to organizational culture, my favorite is “it’s how we do things around here”. This encapsulates how culture is everywhere: it is in the language we use, in how we act and behave with each other, how we reward and penalize people, how we define our strategies and goals, and how we value people and assets.

We consider companies like Facebook, Netflix, Google and Uber as perfect examples of what it means to be data-driven. These companies were born out of data (many say “data is in their DNA”) but do we really appreciate how significant this is culturally?

Think of an organization that during the last 10 years has barely learnt to use ERP and CRM data. It probably uses this kind of data for performance reporting and e.g. informing decisions on sales resource allocation. Humans still make all decisions based on the partial information they have, their personal experience, their own emotions, emotions of people around them, and the fear of making an incorrect decision.

In the analytic maturity continuum, this is typically referred to as practicing “descriptive analytics”. The contrast to the fully automated, non-emotional and purely algorithm-based decision making of organizations born in the data era is significant (Gartner would call to this stage “prescriptive analytics”). “Old” organizations cannot ignore this cultural gap when aspiring for becoming data-driven.

Cultural Traits to Aspire For

What does good data culture look like and how does one get there? [Spoiler alert: There is no single right way, but there are traits to aspire to]

One of the most obvious is consistent language. Language is an essential part of culture but often overlooked. Two simple examples of “speaking data” consistently:

  1. All employees are able to simply explain the business logic of the recommendation engines in the web store, and
  2. All employees understand which data points are used in the optimization algorithm that adjusts factory climate?

In most cases, improving the data literacy will have a bigger impact than any single new technology deployment, and it only requires people time and effort – no software or hardware required!

Have you considered how common the use of gut feel and personal experience is in your decision-making? Daniel Kahnemann’s book “Thinking Fast and Slow” talks about how biased we are in our decision-making and the book has become a popular book on this subject. If you want to be data-driven, you have to acknowledge and actively mitigate for this human bias. In data-driven organizations, decisions are purely fact-based. Gut feel, instinct and individual’s subjective opinions are not factors in the decision-making.

What happens if there isn’t enough data to support a decision? You assume or make “best judgement calls”? Wrong. Data-driven organizations typically do what innovation organizations do: Test, measure (using data), learn and iterate. This requires a culture of curiosity and being comfortable with not knowing something.

How often do your leaders openly admit to not knowing enough or not having an opinion on a critical matter? The world today is often too complex to fully understand and model. Data-driven organizations accept that the best way to make fact-based decisions is to test first and then use data to drive the decisions on what happens next.

Culture Is Also About Values, Ethics and Choices

While “old organizations” have a long way to transform their cultural habits to become data-driven, they should also maintain their integrity and ensure their data-drive is not in conflict with their values. For example, organizations that value their customers over everything else should have no ethical dilemmas in deciding what customer data to collect and not collect. These ethical choices can make them stand out in this current era of digital and data confusion, and can be a significant competitive edge over organizations for whom data is an asset and customers are (only) data suppliers.

A strategy is not just a technology deployment plan, so make sure your strategy includes elements required for a sustainable change.

Culture should be among the first things to focus on in becoming data-driven:

  • Develop a consistent data language and make everyone able to speak it
  • Examine your decision-making and be prepared to let machines take over
  • If you don’t know something, collect data first, then decide
  • Build on the strength of your existing cultural values and ethics

Like Mordor, culture does not sleep. One should not just simply create a data strategy without considering the culture of the organization. Don’t let your data or technology strategy become just another delicious breakfast for your organizational culture.

At Solita we combine deep expertise in creating business and other value out of data, data technology and organizational culture. We can help you across the whole spectrum of becoming data-driven. We call it Datanomics.

[1] Research companies like Gartner widely talk about the low success rate of analytics programs/project. The numbers range between 50% and 80%. You can also simply google search for “why analytics programs fail” and review the relevant hits. In March 2018 Gartner’s presented their CDO Survey results, which showed that “Cultural Issues” were by far the biggest roadblocks to success (in analytics). We have also seen this in our client projects.