Far too few businesses are using big data to support their decision making. In January 2018, a senior analyst with Forbes revealed that business competitors are using data to come after your customers. The Harvard Business Review surveyed Fortune 1000 business executives and found the most common reasons for using big data were to decrease expenses, improve operational efficiency, make more informed decisions, and increase revenue. And 80% say their investments in big data are successful.
Basing business decisions on big data is great—except when it’s not. Basing high stakes decisions on poor-quality research is a recipe for financial disaster.
Here are some of the most common big data errors we see:
- Misleading statistics. You may remember the advertisements claiming 80% of all dentists recommend Colgate toothpaste, leading the consumer to believe 20% of the dentists recommended different brands. The truth, though, was that when the dentists were surveyed about the toothpastes they recommended, they were allowed to identify all of the brands of toothpaste they would recommend; other brands could have been equally as or more popular than Colgate.
- Failure to test the survey questions. It’s easy to create a set of survey questions and send them out through SurveyMonkey or Qualtrics. But if you haven’t pretested and piloted the questions, you can end up with questions that make sense to you—but not to the person taking your survey. Recently, we were asked to complete a survey about our spending on wine purchases made at wineries. Unfortunately, it was unclear whether our spending was to include—or exclude—wine purchased from the winery as part of a wine club membership. Had the survey been pilot-tested, this flaw would have been quickly identified and corrected before deployment.
- Biased interpretation of findings. It matters who does the analysis of survey responses to open-ended questions. This is especially true when a survey is conducted in-house because it is difficult for staff to separate themselves from the data. Unless your team has a staff member specifically trained in eliminating bias, it’s better to outsource your research.
- Lack of candor from survey participants. Your customers generally do not want to hurt your feelings. They are not going to tell you directly your annual customer appreciation event is a dud. This is especially the case where staffing is concerned—and even more so if the staff person administering the survey is also the source of dissatisfaction.
- Failure to collect data. Fewer than half of all businesses collect data at all. In retail businesses, many do not have any idea how many customers come through the door each day, which means the average sales per customer is also an unknown. You can get a ballpark idea using a people-counting electronic system. Sure, the UPS or FedEx carrier may walk through each day, as may staff, but that number will be fairly consistent and you will have a tangible way to measure growth in the number of customers coming through the door. And you’ll know which staff are doing the best job selling your product and which may need additional training—or a new line of work.
The Capiche team possesses more than 20 years of qualitative and quantitative research experience. We understand the importance of valid survey tools and test them thoroughly before deployment. Our most popular surveys have been used thousands of times. Your customers can respond honestly, and we can analyze the survey data without bias because we are a third party. Our only interest is helping you achieve greater success. And we can generally deploy a survey and have findings back to you in less than 45 days. Call us at 541.601.0114, email email@example.com, or use our Contact form today to learn how we can help you leverage quality big data to grow your business now.
Leave a Reply