There are benefits of working in Data Science. We get to use big words (“Is that plot showing some heteroskedastic tendencies?”), correct analyses from MBAs (“You want a median there Liam, not a mean”), and adjudicate matters of statistical significance. Net/net, we’re kind of a big deal.
That’s why we felt we had arrived when we were asked to help with Oracle Data Cloud’s involvement in the Cannes Lions Festival this June. But our team’s experience had nothing to do with sea breezes nor Beef Bourguignon! We were asked to create personalized content tracks for Cannes attendees based on their characteristics to enhance the Festival experience. And as we crunched the numbers, we found some interesting results we want to share:
Cannes Delegates Are a Radical Bunch
If any industry conference showcases creativity, challenges to norms and thinking “outside the box,” it is Cannes Lions. So while I expected some difference of opinions among the delegates, I still wasn’t prepared for the amount of variability we observed among delegates after fielding a brief survey asking the level of agreement with statements like:
- If something can’t be measured, it’s not important.
- Advertising fundamentals from the past are still true today.
- Advertisers spend too much on branding campaigns.
- The process is more important than the outcome.
From our results, 43% of responses were either “strongly disagree” or “strongly agree,” but surprisingly, within each question, there were ample quantities of both.
The most polarizing statement (as measured by maximum standard deviation, due to response bifurcation) was, “If something can’t be measured, its not that important.” The average score was 2.08 (tended to “disagree”) though around 20% actually chose “Strongly Agree” or “Agree.”
Interactions among the responses were also interesting. The first statement (“If something can’t be measured, it’s not important”) was negatively correlated with responses from the second (“Advertising fundamentals from the past are still true today”).
In other words, delegates who agree that, “If something can’t be measured, its not that important,” tend to think that advertising fundamentals from the past are NOT true today. These attendees might be considered the “data disruption delegates” of Cannes. If only all the clusters were as alliteratively impressive!
Given delegate responses, we classified Cannes attendees into six clusters.
We acknowledge that all Cannes delegates are unique, but we proceeded to split delegates into distinct groups according to their survey responses. Using a machine learning procedure called k-means, we allowed patterns in the data to emerge and then grouped delegates who shared similar response patterns. Six distinct clusters emerged, each a unique constituency with generally shared beliefs and interests.
For example, one of the clusters of veteran Cannes attendees generally believe advertising fundamentals of the past are true today and advertisers spend an appropriate amount on branding.
A second group, about half the size of the first and composed mostly of “first time attendees,” think advertisers spend too much on branding, that ideas have to be measured to matter and advertising fundamentals of the past don’t matter in today’s world.
Knowing these differences a priori can be very useful when helping attendees navigate an event as extensive as Cannes.
Finally, we matched content with each of the clusters to build a more customized and relevant experience in Cannes.
Each cluster was provided a tailored itinerary with content recommendations that matched the characteristics of the clusters (as judged by the Cannes content team).
This pairing helped reduce the otherwise overwhelming amount of content available during Cannes and highlight relevant options and perhaps a few “against the grain” sessions to challenge the delegates within the like-minded clusters.
And even though we’re not directly experiencing the action on the ground in France, our team is happy to have helped attendees get the most out of their Cannes Lions experience.
Photo: Monkey Business Images/Shutterstock