We partnered with Mars, Incorporated and their Marketing Laboratory at the Ehrenberg-Bass Institute for Marketing Science to investigate what drives video advertising success. Our research study established that our emotion measurement technology can distinguish between ads with high and low sales lift with 75% accuracy across our dataset.
https://www.realeyesit.com/sales-validation
3. *Source: Prof Gerald Zaltman, Harvard Business School,
'How Customers Think’ (p40.), 2011
Over 90% of human
behaviour is driven by
emotions*
Using webcams and the latest
computer vision and machine-
learning technologies, we measure
how people feel as they watch video
content online.
4. $9.9bn
Digital spend on video
content in US 2016.
Video is growing. Fast.
75m
People in US watch
online videos everyday
3kTotal YouTube channels
owned by Top 100 brands
0
5
10
15
20
25
30
2014 2015 2016 2017 2018 2019 2020
Spend $bn
Vimeo
eMarketer
Pixability.com
6. “We need to avoid the
‘crap trap’ of content”
– Marc Pritchard, CMO of P&G
“Great Creative is
kryptonite to ad blocking”
– Gayle Fuguitt, Chief of Customer Insight
and Innovation Foursquare
"Empathy is the superpower of
building brands right now.”
- Frank Cooper: CMO, Buzzfeed
"Most of the ads today
deserve to be blocked"
– Jeremy (Yoram) Wind, Professor of
Marketing, The Wharton School of the
University of Pennsylvania
“Emotional data will eventually
become much more important”
– Tracey Follows, CSO and IO, The Future Laboratory.
“Emotion is the main profit driver
for brands. Creativity matters”
– Les Binet, Head of Effectiveness Adam & Eve DDB
7. “Good” starts
with the creative.
Creative is the most important,
and most overlooked driver of
marketing success.
8. Testing times
for video
Brands go to great lengths to
gain audience insight to video,
but none of them are linked to
real-world outcomes.
Surveys
Focus
Groups
Dial
Testing
EEG
Heart
Rate
Click
Skip
Implicit
Reaction
10. We partnered with Mars,
Incorporated to investigate
what drives video
advertising success.
The study was designed in
collaboration with the Mars
Marketing Laboratory at the
Ehrenberg-Bass Institute for
Marketing Science.
11. Objective/
Identify ads with high
sales potential, using
emotion measurement
via facial coding
through webcams
Essential for Mars,
Incorporated to
understand how to
create successful videos
Verify whether a video
hits the mark before
spending on media
12. Mars, Incorporated provided single-
source sales data indicating high or low
sales impact for each of the 149 ads
Realeyes collected emotion data on
each of the ads
Created a predictive model to validate
which emotional patterns matter
Methodology/
Using World’s largest
emotion database
linked to sales
13.
14. • Model differentiates between ads with
high and low sales lift impact.
• Rigorously validated with advanced
statistical methods.
• Robust model performance across
product categories and countries
indicates the model may be universal
across product verticals.
4 Categories 6 Countries
Results
75% accuracy based
on emotions and
behavioural signals
15. Results
Example variable:
Keep happiness
growing
• Variable 1: time of happiness growing
vs time of happiness falling
• Successful ads grow happiness for
longer than having it falling,
unsuccessful less so
• Absolute level of happiness matters
less than the dynamics of it
developing over time
16. Conclusion
Facial coding provided a model that
enabled to separate good ads
(proven in-market sales) from bad
(zero/low sales) with 75% accuracy
Could guide smarter media spend
decisions, at scale
Next: explore how the key success
signals can inform creative processes
17. “People will forget what you’ve said,
people will forget what you did,
but people will never forget
how you made them feel.”
– Maya Angelou