Mauricio Ferreira leads the Advanced Marketing Sciences group at Hypothesis (the BrainTrust). Contact him at firstname.lastname@example.org.
Just the other day, discussing a plan with a client, the topic of a Derived Importance came up. The discussion made me realize that despite being a widely used technique in marketing research, derived importance is often misunderstood, especially the interpretation of its results. Here are answers to 3 of the most frequently asked questions about derived importance we receive from clients.
"What is a Derived Importance analysis?"
Derived importance is essentially a statistical method used to understand what “drives” a variable of interest. For example, we may use it to understand what elements of a message drives interest in an ad or what service features can lead to satisfaction. The analysis helps marketers know what to prioritize and emphasize in product development, service improvements, or messaging.
The term "derived" means that the importance is extracted from the statistical relationships between metrics rather than by asking consumers directly. The derived measure of importance represents the partial contribution each driver makes in explaining or predicting the outcome.
"Why do we derive and not just ask people?"
We may be tempted to simply ask consumers directly what is important to them, but this approach often produces undifferentiated results. That's because respondents say everything is important (e.g., price is very important, quality is very important, etc.). Another problem is that consumers may place high importance on “price-of-entry” variables which is not useful when interpreting results. For example, not crashing is obviously very important for air travelers, but it’s probably not a useful marketing message for an airliner. Also, respondents might feel socially compelled to cite certain variables as important (e.g., safety) and others as less important (e.g., brand image), but we know that in reality, brand image might be a bigger determinant of choice in certain categories, like automobiles.
"Why isn't 24-hour customer service at the top of Derived Importance?"
Sometimes, a variable that seems like it should be important, comes out low on the results of a Derived Importance analysis. “How could this be? Are you saying 24-Hour Customer Service is not important? But, our customers always mention this in focus groups!”
That’s part of the power of a derived importance analysis. It’s not about identifying important variables per se, but rather variables that will move the needle. For example, 24-hour customer service may be important to consumers, but if all relevant brands offer it, then all brands would score high on it. In such a case, there will be no variation in responses between the attribute and outcome, and 24 hour customer service would actually fall low on the Derived Importance analysis – as it should, because emphasizing something that everyone does will not differentiate your brand.
(As a cautionary note, when we find an attribute rated low on the Derived Importance analysis, it doesn’t mean that it should be overlooked. It’s still important for an airliner to not crash!)
If you have more questions about a Derived Importance analysis, feel free to contact me directly at BrainTrust@hypothesisgroup.com. Also, if you have a question on another topic, please let me know!