• Paul Longhenry

Discovering a "Known Unknown" in the Digital Advertising World

Updated: Mar 5

There are a near infinite number of variables that can trigger consumer engagement for a given ad campaign, and only so many can be known and tracked. The best platforms, DSPs, and networks actively leverage the largest set possible, identifying those delivering predictive signal, and rapidly replicating success.

- Resonating with customers with a particular purchase history? Behaviorally targeted.

- Driving clicks from guys in their 30s? Demographics optimized.

- Tripling conversion rates across sports sites and apps? Context prioritized.

- 60% of engagement with a particular creative? A:B testing complete.

Yet, after all this rigorous optimization, click and conversion rates usually remain inexplicably inconsistent. There isn’t much else that can be done; there are simply too many unknowns about all of us consumers as individuals. We’re all a lot more complicated than our purchase histories, demographics and content interests, and there is only so much that algorithms can do with these limited data sets.

While only a piece of the remaining puzzle, here at Pinpoint we have been working to uncover what we consider to be one of the larger remaining “known unknowns” that could move the needle: personality. To be clear, we’re not suggesting that personality insights can predict a person’s interest in buying a new dining room table tomorrow, rather, we’re talking about personality-optimized creative.

Intuitively, we understand that specific words, phrases, colors, and images appeal differently to different people. Academic research supports this intuition and demonstrates a strong relationship between an individual’s personality and the words, colors, and images that prompt an individual to engage. One such example is out of Cambridge University, where they analyzed a Hilton target audience and found that engagement rates diverged significantly based upon relative levels of extraversion and agreeableness. The team then crafted an ‘extraverted’ creative (groups of laughing people playing volleyball and barbecuing), and an ‘agreeable’ creative (families enjoying time together on the beach). These creatives were then served to either matched personalities (extroverted people receiving extroverted creative) or mismatched personalities (extroverted people receiving agreeable creative). The results were decisive: creative matched to the targeted personality resulted in clickthrough rates twice the baseline.

While this was just a small test, the emerging field of “Psychometric AI” brings quantitative psychology together with machine learning for much larger audiences, and is able to infer the detailed personality characteristics of anonymized individuals at scale. This enables statistically significant measurement and understanding of how a given creative resonates across distinct personality traits, and points the way to personality-based insights as a strong compliment to today’s state-of-the-art practices (namely DCO, or dynamic creative optimization).

Pinpoint is a pioneer in Psychometric AI, and we’re very focused on its potential in creative optimization. We’ll have much more to share about the known unknown of personality-based optimization in the months to come.