I spent last week at a conference focused on customer research. There were +1,000 researchers in attendance, mostly from Fortune 1000 companies. They were brilliant — discovering insights that could lead to better products, more satisfied customers, and higher profits.
The problem is that their customer research isn't able to produce results. Disappointed, I started to think about how researchers could make more of an impact on the organization.
My first thought was that researchers are not senior executives. Perhaps researchers should give senior executives the responsibility, training, and tools to conduct their own customer research? It's an interesting idea — a lot of vendors are promoting self-service business intelligence — but how feasible is it? If you give executives the power to do research without proper and extensive training, you are going to have an organization led by misinformation.
My second thought was that customer research teams are too small. For example, customer research at E*Trade is carried out by a team of 7 employees — with no help from contractors. At first I was shocked, but I found that this is pretty common. Most Fortune 1000 companies have between 5 and 15 researchers focused on helping the corporation understand their customers, and they are pretty confident in using technology and outside vendors to keep up with the organizations needs — so size doesn't seem to be the problem.
What is needed is a different way to deliver customer research.
As the CEO of Contemporary Analysis, I have a lot of big insights about the future of CAN, predictive analytics, and business — and I know how fun those insights can be! But, I have learned that there are appropriate ways to convey them.
If BIG INSIGHTS are delivered all at once, everyone runs! Even good changes need to be introduced slowly. Researchers need to do the following to make their customer research more impactful:
- Researchers should connect specific actions to goals and employees. Feedback should connect employee action to the whole — companies achieve results when employees move together in a common direction. To do this, researchers must understand the drivers of customer statistics they are looking to improve. This can be accomplished by building description models and/or predictive models.
- Individual workers need personalized customer statistics. Net Statistics are useful, but they don't inform the individual worker about their work. It is interesting to know what the whole group is doing, but individual employees need to be informed about their performance.
- Individualized statistics must be relevant. Feedback should be delivered during activity or within memory of an activity. Imagine driving a car: every month you receive a report of every trip you have taken, your average speed, and a histogram of the frequency of the speeds you traveled. This report might be really interesting, but doesn't help increase your gas mileage or avoid speeding tickets.
- Teach — and train — to the action, not the statistic. Don't explain that someone's statistics are off; instead teach them what actions they can do to change their statistics. Researchers should encourage managers to do role playing both planned and random to reinforce the training. Employees can't change the metrics, but they can change their actions with customers.