Piaggio’s Question: What combination of attributes in our ads will positively impact the perception of our brand? Our survey design helped the scooter-maker identify the right ads for distinct consumers.
Piaggio, the makers of Moto Guzzi and Vespa, needed to identify the best online ad but testing Piaggio’s desired combinations using traditional A/B testing was too difficult.
What We Did
In regular A/B testing, clients can only test simple comparisons. Civis Analytics used conjoint analysis to test the full range of options Piaggio’s creative team developed ““ from color of scooter to message and background ““ against the full set of their desired consumers.
Results that Matter
Instead of delivering one winning ad, Civis was also able to assign winning content by consumer segment. Piaggio implemented a data-driven ad campaign that told a story driven by consumers instead of gut feeling.
Good survey design leads to impactful conclusions
For the uninitiated, conjoint analysis looks something like this:
In practice, we were able to test, via surveys, the following advertisements with both scooter owners and prospective scooter owners.
Good survey design leads to impactful conclusions. Not only did our survey give Piaggio a winning ad “overall,” but we were also able to assign winning content by consumer segment:
While affordability was the most stable feature, appealing to prospective owners (“new leads”) and current owners alike, this matrix of features allowed Piaggio to deliver the best ads to consumers based on preference.
Beyond allowing us to test these combinations, our survey design linked stated ad preference with an action: in this case, “test drive a Vespa” at a local dealership. This was our answer to a possible discrepancy between stated preference and revealed preference. Piaggio was able to trace customer preference for an online ad to their most desired offline outcome: a customer in one of their brick-and-mortar stores.
Written by: Caroline Grey
The post Piaggio: The Accurate Way to a Story Worth Telling appeared first on Civis Analytics.