Introduction
Data analysts often struggle to provide actionable insights for effective business decisions. The key lies in ensuring that data analytics is decision-driven rather than merely data-driven.
The Pitfalls of Data-Driven Decision-Making
While data-driven decision-making is a buzzword in management, many executives express disappointment in the results of their data analytics initiatives. The article argues that the flaw lies in the traditional approach of finding a purpose for the data at hand, leading to irrelevant questions and the reinforcement of preexisting beliefs.
‘Data-Driven’ Often Means Answering the Wrong Question
Illustrating with RollingBoulder, a media company, the article demonstrates the limitation of the conventional approach. While the company successfully predicts customer churn, it fails to answer a crucial question – the impact of including a gift on a customer’s likelihood to churn. The article contends that data-driven decision-making tends to anchor on available data, often addressing the wrong question.
‘Data-Driven’ Often Means Reinforcing Preexisting Beliefs
Examining the case of Gwenn & Jenny’s, an ice cream vendor, the article exposes the pitfalls of accepting data that aligns with preexisting beliefs. The three-step process used by Twitter to measure the impact of advertising is criticized for exaggerating the influence of social media on sales, reinforcing clients’ beliefs rather than providing objective insights.
Moving to Decision-Driven Data Analytics
To overcome the limitations of traditional data-driven approaches, the article proposes a shift to decision-driven data analytics. This involves identifying key business decisions, selecting alternative courses of action, determining necessary data, and ultimately choosing the best course of action.
Steps Towards Decision-Driven Data Analytics
Identify Alternative Courses of Action:
Decision makers are urged to think “wide then narrow,” considering multiple courses of action before narrowing down choices.
The example of RollingBoulder is revisited to emphasize the importance of generating a high-quality and feasible consideration set.
Determine Necessary Data:
Decision makers and data scientists collaborate to establish criteria for ranking alternative courses of action.
The goal of data analytics is highlighted – turning unknowns into knowns to enhance the objectivity of decision-making.
Select the Best Course of Action:
If the first two steps are executed well, data analytics should reveal the best course of action.
The article cites the example of RollingBoulder employing a randomized controlled trial to understand the impact of adding a gift to a customer’s renewal letter.
Conclusion:
Decision-driven data analytics emphasizes the importance of asking questions and managerial judgment. By focusing on actionable insights tied to key decisions, this approach ensures that analytics initiatives challenge executives’ beliefs, providing more meaningful contributions to the decision-making process.