Who wouldn’t love love love having data to support decision-making?
Interviews with two dozen people, mostly marketing managers and executives, on the subject of data-driven decision making, reveal deep-seated ambivalence.
There are valid reasons to reject the value of data for decisions. Data quality is a pain point for everyone, and quality may be so poor that analyses are dangerous. Startups change so rapidly that analysis of even last quarter’s data is useless. Medical privacy argues against collecting or even using visitor data.
People have sources and methods, and don’t believe the cost/benefit of trying to improve these. They are comfortable using business intelligence reporting and web analytics, recognizing the imperfection and incompleteness. People feel they understand enough.
Spreadsheets of downloaded or hand-collected data are the common approach to answering new questions. Data is incomplete, analyses are limited, and collaboration is a manual process.
But it’s a familiar process, with very familiar flaws. Who knows what flaws will come from other approaches?
Bottom line: Attitudes and habits are obstacles to adoption of data science technologies and methods.
On Target 