Turning A Data Science Vanity Project Into Something Useful: Thor's Abs
As I mentioned in the last post, Thor's Abs is a vanity project, and that's dangerous for any Data Scientist. Most low data maturity businesses run on metrics and KPIs that are disconnected from business goals. This disconnect is the root cause of so many different problems across the company, and vanity projects are just one of them.
For Data Scientists, vanity projects create a trap. The business has a strong understanding of its goal but a weak understanding of what metrics cause the company to reach that goal. As a result, project selection is biased by strongly held beliefs about what metrics matter and what projects move those metrics.
That means I can deliver a successful project, and the project's success metrics might not move. If the success metrics move, the business may not see the expected progress towards its primary goal. A vanity project is a setup to fail. Before I succeed with this or a future project, I must change the way the business makes decisions.
The most significant barriers to change are strongly held beliefs. People create heuristics based on strongly held beliefs to reduce the cognitive load of complex decisions. Decision makers hate the introduction of data because it increases the cognitive load of decision making, or at least their perception of cognitive load, by dragging them out of comfortable heuristics.
That's why getting senior leaders to adopt data driven methods is a challenge. Having data in front of them and getting them to use that data to improve decision quality are two different phases. Users won't adopt anything that makes their lives harder, so I need to understand some decision making fundamentals to set the decision support system up for success.