A friend works at a company that is very data driven. The beliefs of this company are laid down in a nice booklet for all staff. Belief number 6: “We do not rely on guesswork. We base our decisions on data. We test our ideas before we implement them”. This is a great practice that deserves wider application.
In advertising the practice is called A/B testing. Two advertisements are developed. One with the color red, and the other with the color green. Both advertisements are run, and the reactions are counted. The winning advertisement is used. With web advertisements it is easy to use A/Z testing: make advertisements with all kind of different colors, and see what happens. Women might like the green one, and men the blue one. You can then use the green advertisements in media targeted at women, and the blue advertisements in the media targeted at men.
HR could use A/B testing a lot more. Some examples:
- Of course the easiest area is recruitment. Develop two different campaigns, run them both and see what works best.
- Talent development. Give two top potentials similar assignments, and see after six months or a year who performed best.
- Innovation. Spilt the team in two, A and B. Give both teams the same assignment and see which team comes up with the best idea.
- Training. Develop two different training modules to train a certain skill (e.g. project management). One on-line version and a classroom version. Split the target group in two, and see who show the biggest increase in skills after the training. And keep tracking: who delivers the best results long-term?
- Incentives. The design of effective incentive systems is not easy and generally not very data driven. A/B test your design. Split the team in two. Group A gets your very sophisticated incentive system. Group B gets no incentives. You might want to add a group C, where the incentives are distributed discretionary at the end of the period.
- CEO-succession. Candidate A gets Division X and candidate B gets division Y. The candidate whose division performs best gets the job.
I am sure (and I hope) you can come up with many more examples. Measurement can be difficult, as well as creating comparable conditions for the A and the B group. Some difficulties cannot be completely overcome, but I am convinced our profession can benefit from doing a lot more A/B testing.