0

Analytical tools begin with psychology

Analytical tools

Feels like companies these days spend a lot of money on analytical tools, either in the form of software suites or highly-compensated data scientist-type people. Makes sense, of course: data is the “new oil” (in reality AI is the new oil but let’s gloss that over for now), and companies need a way to compete. Many executives rose up claiming to understand data (i.e. “the numbers”) but really don’t, so having someone on the team who “gets data” is an increasingly important skill.

Problem: most people aren’t doing it right.

What’s wrong?

A few things. At too many places, the focus around data and analytical tools is:

  • Collect the data
  • Tell everyone you’re data-driven

Data/analytics is absolutely useless unless it’s tied to better decision-making. Oftentimes, ’tis not. For decision-making to improve, the data team needs to have someone on it who can explain the data. At most places I’ve seen, the data team is constructed around people who can (a) collect the data or (b) put it into cool-looking charts. There aren’t many people who can take the information and present it around terms that the check-writers would understand. We need more data translators.

But there’s a bigger, first problem we’re missing.

The psychology of analytical tools

People smarter than me, i.e. Stephen Dubner, have pointed this out too. Work is closely tied to self-worth and the need for relevance. If the data you collect contrasts with your ability to do those things (i.e. if it says the opposite of your opinion), you probably won’t trust the data. My last full-time job was this every day. We’d run all these Google Analytics reports, submit them to VPs, and they’d say “No, nice numbers, but we trust our gut.” What the fuck is the point of collecting/running data then?




 

There’s also this, from a new article on Harvard Business Review:

On the other hand, analysts who are too deeply embedded in business functions tend to be biased toward the status quo or leadership’s thinking.

That too. Let’s set this up simply:

  • You have data.
  • You have a boss.

If the data says one thing and the boss says another, well, you should go with the data. Most would go with the boss, though. The data can’t promote you. (Well, it could if your organization was really forward-thinking.)

How could we get better at analytical tools?

A few ideas:

  • Stop believing software is going to save us
  • Hire the right people, i.e.
  • … someone to collect the data…
  • … a person to scrub the data …
  • … a third person to analyze the data …
  • … and someone who presents the data to executives
  • Root decision-making in data and analytics
  • Don’t let certain stakeholders get out of it by “trusting their gut” or “knowing the industry”
  • Make everyone responsible for analytics at their level
  • Don’t lip service the concept
  • If you know you have execs who just want to pound their chests, don’t even try anything around analytical tools
  • Care

What else would you add on analytical tools?

Ted Bauer

Reply If You'd Like