The psychology of data-driven decision-making


Tweet about this on TwitterShare on FacebookShare on LinkedIn

A railway track switch

How often do you find yourself caught sitting on the fence? Why is it often difficult to make simple decisions such as where to eat or what movie to watch? Decision-making is a mental process that can be influenced by a number of factors such as reason, bias, emotions, or past experiences.

Data-driven business professionals who are tasked with making quantifiable decisions know that decision-making goes beyond following gut instincts. It involves rational analysis of data, decisions and actions. Today we’ll be taking a look at the psychology of decision-making and some tactics to help your organization make high-quality decisions.

Psychology Today suggests there is a time and a place for instinct driven decision-making in their insightful article Gut Almighty. Numerous studies have led them to conclude that gut instincts are particularly helpful in games such as trivia or tests, where overthinking won’t necessarily lead you to the right answer. For example, in a multiple choice question, one answer may stick out more than the others because of prior knowledge exposure.

Emotion-based gut decisions, such as purchasing a home or choosing to marry, are often made by relying on past experiences. We go off of hunches, especially when getting to know others. It’s simply unnatural to try to crunch numbers when deciding whether to pursue a relationship or where to plant your roots.  

However, in business; hunches, emotional instincts, or bias can be detrimental to arriving at a solid decision. They will not necessarily add value to the process, so decisions aimed at optimizing performance and meeting strategic objectives should not be made on the basis that they ‘feel right’.

In a study conducted at Berkeley, defective decision making was found to be the result of:

  • Poor information analysis
  • Lack of weighing alternatives
  • Failure to deliberate risks
  • Quick judgment call

The study also goes on to explain how humans employ cognitive bias in an attempt to digest large amounts of, or complex, information. There are a number of biases that can come into play including:

The Monte Carlo or Gambler’s fallacy: the erroneous belief that chances of succeeding mature. For example, when tossing a coin, one may believe that if tails is flipped multiple times, then the odds of flipping heads are higher. In reality, however, the outcome is independent of previous results.  Although  marketers do have tangible results to strategize around, some continue to do the same thing with hopes that the odds will suddenly shift in their favor.

The overconfidence effect: the belief that one knows more about a subject than they actually do. For example, a CMO may forecast that their upcoming display ads will have a 20% click through rate when in all reality they barely tip 3%.

Framing alternatives: how we frame our options or information available has an impact on how it is processed and acted upon. It is found that we are more prone to act when the result will be negative. For instance, if forecasting shows us that an investment of 10% of our marketing budget in social channels will lead to improved performance, we may be more successful in gaining approval from the budget owner for this allocation if we frame our proposal as “we will lose the opportunity to increase engagement by 50% if we don’t make the investment”, rather than telling them that “this investment will lead to a 50% increase in engagement.”

Luckily, biases can be avoided when teams are encouraged to explore alternatives and educated not only on the effect of bias, but also on the information at hand. In order to make decision-making as bias free and as objective as possible, there are a number of models designed to aid organizations to collaborate and arrive at high-quality decisions:

RAPID by Bain & Company, or ‘Input, Recommend, Agree, Decide and Perform’, encourages organizations to define roles, strategically evaluate the value of decisions, and act. However, for successful execution of this model, a strong level of commitment is required throughout the organization.

The Lean Analytics Cycle, created by  Alistair Croll and Benjamin Yoskovitz, is designed to put learning at the heart of decision-making. This 4-step process emphasizes the values of acting and testing. This model is primarily oriented towards start-ups, but it can be adapted to fit the needs of larger organizations.

Digital Insight Management, developed by Eric Peterson & Sweetspot, goes beyond decision-making tactics and explores how to leverage your existing investment in digital analytics in order to effectively collaborate and act upon insights. This model encourages organizations to focus on key objectives, define workflows and measure the impact of decisions made.

Collaborative or group decision-making, however, often presents challenges of its own that can impede successfully arriving at an optimized decision. The three models above each highlight the need for internal buy-in for decision-making processes to be successful. The evolving role of data in decision-making report by The Economist Intelligence Unit also confirms that it all starts with company culture.

The study found that companies who do not use big data as a valuable resource in their decision-making were less successful, whereas those who combine it with human expertise were more likely to make informed decisions. Furthermore, inadequate reports were often to blame for suboptimal decisions where data was inaccurate or inconsistent.

Building a culture around data-driven decisions presents organizations with an immense amount of opportunities to analyze key success metrics, weigh the impact of their decisions, and optimize their capacity to learn and grow. In order to get started, it is necessary for organizations to utilize analytics and reporting to ensure that reliable information is delivered to executives so that they can access it freely and act upon it whenever needed.

Exploring data-driven decision-making & storytelling at the DAA New York Symposium

Next Monday we’ll be at the DAA New York Symposium 2016. Apart from looking forward to having a chance to catch up with members of the NYC Measure community, we are eager to learn from the speakers, including professionals from Huffington Post, IBM and NYU, who will share their experiences and insights. The sessions, covering strategies for increasing analytics engagement; from storytelling to data journalism, each look more interesting than the last.

Jim Thompson of Billboard Magazine and The Hollywood Reporter will be there to speak on “Solving for the Problem of Choice”. We’ve previously shared our ideas on how to overcome decision paralysis, but we’re eager to further explore the psychology of data-driven decision-making there.

Kristen Sosuluski of NYU will also speak on Engaged Storytelling with Information Visualization, a topic close to our hearts with the recent release of our White Paper The Path to Data-Driven Storytelling and Actionable Insights.

If you’re attending the DAA Symposium we’d love to meet you and hear your thoughts on the challenges behind data-driven decision-making. If not, please share your thoughts below. Have you successfully implemented a decision-making model in your organization?

Tweet about this on TwitterShare on FacebookShare on LinkedIn

Holly McKendry

Sweetspot Marketing Director. Wakeboarder & travel enthusiast. Communication Studies graduate of Texas State University, San Marcos.

Add a comment

Try Sweetspot today!

Not Another Dashboard.