You’d be hard pressed to find a business leader who doesn’t agree to the conventional wisdom that good businesses manage by data. Indeed, data remains a perennially hot topic among biz-dev consultants — how to collect it, how to develop “metrics,” how to package it for senior executives, etc.
My own experience suggests three major faults with the current data-driven management paradigm.
First, the stats we use are too superficial. People don’t like admitting that they don’t know something, so most analysis resides at the intellectual level of the lowest (reasonable) common denominator. As such, you’ll see at lot of companies with run charts, fewer with control charts, and almost none (outside of specific science-focused industries) with complex regressions or correlations. The net result is that these companies are adept at picking the low-hanging fruit but are functionally blind about the juicer pickings further up the tree. Stats 101 will only get you so far.
Second, we don’t trust our analysts. Either managers demand reports and data and then do their own interpretations, or we let the stats people perform conclusions but we keep them out of meetings and away from the levers of influence. It’s rare to have a high-performing analyst who is known for being an analyst be present as a matter of course at leadership-level meetings.
Third, we use data as a lever for internal politicking. We demand reports that are never consumed, or we keep bickering about the data or the metrics or the benchmarks as a way of delaying, shunning or transferring responsibility for poor performance. Likewise, controlling the analysts leads to controlling the spin.
So, what will a high-performing company that really does live by its stats look like? A few ideas jump out:
- The data people will work in their own shop and be largely free of mid-management interference into scope and methods. The “every manager has his own analyst” school of thought doesn’t advance the overall organization’s effectiveness.
- The company as a whole will provide various in-services or other forms of education to help low- and mid-level managers improve their statistical competence.
- The company will emphasize information instead of data, letting the analysts perform complex statistics and releasing their plain-English results to others — and the audience will trust that the analysis is correct and unbiased.
Knowledge is power, and data yields knowledge. For a company to thrive in a data-driven management climate, let your independent analysts release information (not data) to inform strategic decision-making.