August 15, 2018
Market research that’s properly executed and reported is a great tool. But, unfortunately, a lot of research is pure bunk. And that’s seriously dangerous. Especially if the results will inform a major marketing decision about a hospital or medical device If you want to learn about your staff’s favorite snacks, by all means, fire up a poll on Facebook. Otherwise, keep an eye out for these pitfalls:
Management Objectives MIA. Every research project needs a management objective. It’s a clear, simple declaration of what decisions you would like to make based on the research. Start by determining what features your doctors, patients and customers value the most. If you didn’t include research in the development of your program, prepare to be surprised.
We launched a sophisticated neurosurgical microscope and discovered that the key to the sale was not the cool neuro-navigation feature, but one-touch balancing. Consumers are unpredictable, so assume nothing, and get reliable information. Decisions may include adding or eliminating features, discontinuing the product, or modifying style. The management objective is also key to C-Level funding and action after the study.
Fake Focus Groups. You know a focus group is fake if you’re conducting it yourself; the focus group members know who’s sponsoring the research, or the participants don’t represent your target market. For honest feedback, hire a trained moderator and book a facility that doesn’t give away your hospital’s identity. A prime example is trying to elicit patient satisfaction feedback by inviting patients to take an “anonymous” survey on your website. Any hint of surveillance will skew your results, or discourage participation at all.
You know a focus group is fake if you’re conducting it yourself.
The Math Trap. Quantitative research is meaningless unless reported with a confidence level. The confidence level is the percentage of the population that you can expect to duplicate any given result. For example, you’ve enlisted 100 male and 100 doctors car to survey colors for exam rooms. You discover that 10% of men prefer their rooms in a blown out rust color. Chill, it’s not a new market niche.
Without a confidence level, you might infer that 10% of men are rust aficionados. That would be ever so wrong, because the sample size is small, which leads to a very low confidence level. In general, confidence level goes up the larger the sample and plunges when it’s smaller. Without a confidence level, your information is cooked.
Bad Recruit. The usefulness of quantitative or qualitative data depends on how accurately your focus group members or survey sample represents your customers. For example, you don’t want to talk to Florida residents about snow tires.
The best respondents are those that are randomly selected from the population you want to research. Not your mailing list, extended family, or former sorority sisters. Define your
Make sure your study is wide enough, so you consider competitors of your product or service, as well as meaningful samples from various age groups. Generally, the more detailed the recruiting profile, the more expensive the recruit. But it’s worth it if you value confidence in your results.
DIY Questionnaire Design. The answers you get depend on how you ask the question. “Do you agree that Michael Jordan is the greatest basketball player ever?” [ ] Yes [ ] No. That’s a perfect example of a biased question.
It should be crafted like this. “Which, if any, of the following basketball players, would you consider the greatest ever?
None of the above
It’s a small thing, but it can seriously skew results.
You need a certain type of mojo to write and design an effective survey. There are more ways to create bias in asking questions than combinations of deli meats to put on a sandwich. Questions and answers need to be rotated. The right types of rating scales need to be used, and so on. Again, if you want serious questions answered, hire a professional.
Tell Me What I Want to Hear. We already showed you how to play with confidence levels to manipulate results, possibly to protect a pet project, but usually it happens at a lower level of consciousness. You and your boss hope for a certain result, and it expresses itself in a bias when you ask a question. It’s more than enough to invalidate a study.
Hire a competent research firm, especially if a program or people’s jobs hang in the balance.
- Avoid DIY focus groups. They’ll tell you what you want to hear.
- Recruit the right people.
- Insist on confidence level reporting in your quantitative research.
- Filter out your personal bias by removing yourself from the tactical
- phase of research.