Research mistakes are preventable
Tom Dougherty, CEO – Stealing Share
25 April 2017
Research mistakes. No excuses.
Making research mistakes is easier than ever. And these blunders can mangle and mislead brands. Especially brands that are looking for opportunities to rebrand. It’s best not to guess. It’s always better to know.
Thinking you know what you know is a bad place to begin. Research mistakes are plentiful when you don’t challenge your own thinking.
Here is an example of what I mean. For years, CATI studies (Computer Assisted Telephone Surveys) were the gold standard in market research. Nothing substituted for this methodology. But CATI research is expensive.
Not everything is what it seems
Today, CATI studies are harder to field. And not just because they are expensive. How many of you have a land-line telephone? That group of dinosaurs shrinks every day. We don’t need them any longer.
Many researchers tell you they compensate for that shrinking population by including cell phone numbers. A good idea in theory.
But, tell me, how often do you answer a call on your cell phone from an unknown caller? Worse still, an unknown caller from out of the area?
Research mistakes happen when we make assumptions
Methodology matters when it comes to projectable research. Using flawed methodology clouds results. Your research study can’t be self-selecting. The goal is to eliminate bias and get a fair representation of the target.
So, randomization equals great research. And being double-blinded. But it also relies on what researchers call incidence. Incidence is the percentage of randomly selected subjects that complete the study. The lower the incidence, the higher the margin of error.
“Research methodology is changing. It is keeping up with changes in demographics and privacy. What worked three years ago is outdated today.”
This make sense. Common sense.
Think about this for a moment. If over half of the possible target audience refuses to complete the study, there may be a bias for those that do. Once a large percentage of respondents refuse to take the study your randomized sampling is anything but random.
To compensate for this, we pay many respondents an incentive to complete the study. This used to work well. The researcher tells you upfront that you are paid for your thoughts.
It fails if you can’t get through
But you’re sunk if your target audience does not use a land-line or won’t answer a caller they don’t recognize. No amount of incentive can compensate for an inability to break through the barriers of availability.
On a recent CATI study, 75% of population were over the age of 65. Tell you anything?
Panel studies are now the GOLD Standard
You can predict the success and value in panel studies by the confidence you have in the panel itself. We have found them more accurate than CATI studies in many categories. And we are quickly finding them preferable in general population studies too.
It’s a new world in research. Research mistakes occur when we don’t challenge our own assumptions.
Too much customer service Tom Dougherty, CEO - Stealing Share 21 June 2018 Too much customer service is a real thing I’ve said this before, but I gotta say it again. Too much customer service drives me bat-shit crazy. Not surprisingly, Best Buy kills me with this....
Dow Jones ousts GE Tom Dougherty, CEO - Stealing Share 20 June 2018 Dow Jones kicks out GE. Now what? In 1896, the stock market adds GE as one of Dow Industrial Average’s first companies. As of June 26th of this year, it will no longer be part of the DJIA. I typically...
Tim Cook is leading the way Tom Dougherty, CEO - Stealing Share 19 June 2018 Apple CEO Tim Cook Tim Cook. Apple Act 2. Like everyone else, I worried about Apple’s succession plan. I was a died-in-the-wool member of the Steve Jobs cult. I was not sure what to make of...