1. EKSTRA BLADET NETSALG
D
Predictive Behavioral Targeting
at ekstrabladet.dk
ekstrabladet dk
2. Introduction
EKSTRA BLAD
Maria Voigt
Project Manager – ekstrabladet.dk
DET NETSALG
Marketing Research Department
Predictive Targeting Evaluation Project
Advertising Method Expert
Stephan Noller
Managing Director – nugg.ad
Founded nugg.ad in 2006
Profound experience in data analysis
at TNS Group (Emnid, Infratest)
at industry association AGOF
3. Introduction to ekstrabladet.dk
EKSTRA BLAD
Part of the media concern
JP/Politikens Hus
DET NETSALG
Largest news site in Denmark
The site:
Breaking news
User profile
In comparison Ekstra Bladet is like:
Aftonbladet, Sweden
VG, N
VG Norway
Bild, Germany
The Sun, UK
4. The Danish Market
EKSTRA BLADET NETSALG
D
90% of the Danish population
p p
have internet access.
Danish internet population
4 million real users
3 billion page impressions
5. Challenges
EKSTRA BLAD
Demands for:
DET NETSALG
Better results (more clicks, more sales)
Documented pay-off
Is targeting the solution?
Ekstra Bladet tested two campaigns using
Predictive Behavioral Targeting (PBT) from nugg.ad
6. Predictive Behavioral Targeting (1/3)
EKSTRA BLAD
Jim | Jane | Tom Tim
ekstrabladet.dk/news 1 2 1 1
DET NETSALG
Content Channels ekstrabladet.dk/people 3 1 3 3
(Measured) ekstrabladet.dk/sport 0 2 3 3
ekstrabladet.dk/tech 2 1 2 2
Step 1: Click Measurement (all Users)
7. Predictive Behavioral Targeting (2/3)
EKSTRA BLAD
Jim | Jane | Tom Tim
ekstrabladet.dk/news 1 2 1 1
DET NETSALG
Content Channels ekstrabladet.dk/people 3 1 3 3
(Measured) ekstrabladet.dk/sport 0 2 3 3
ekstrabladet.dk/tech 2 1 2 2
Demography Gender 0 1 0
(Survey) Age 3 3
Product Interests Body Care 2 4 2
(Survey) Food 0 1 2
Psychography (Survey) Trendsetter 0 2 2
Step 2: Online Survey (Sample)
8. Predictive Behavioral Targeting (3/3)
EKSTRA BLAD
Jim | Jane | Tom Tim
ekstrabladet.dk/news 1 2 1 1
DET NETSALG
Content Channels ekstrabladet.dk/people 3 1 3 3
(Measured) ekstrabladet.dk/sport 0 2 3 3
ekstrabladet.dk/tech 2 1 2 2
Demography Gender 0 1 0 0
(Survey) Age 3 3 3
Product Interests Body Care 2 4 2 2
(Survey) Food 0 1 2 2
Psychography (Survey) Trendsetter 0 2 2 2
Step 3a: Search for Tims nearest statistical twin
p
Step 3b: Tim inherits missing values from his “twin” Tom
9. Case Study – FDM Travels
EKSTRA BLAD
Campaign: weekend holidays
Ca pa g ee e d o days
DET NETSALG
Aim: sales/purchases
Creatives:
10. Case Study – FDM Travels
EKSTRA BLAD
300
DET NETSALG
250
Index Click Through Rate
200
R
150 136
no PBT
C
100 PBT
100
50
0
Holidays
Day 1‐2
11. Case Study – FDM Travels
EKSTRA BLAD
300
DET NETSALG
250
Index Click Through Rate
200
R
150 136
no PBT
C
100 100 PBT
100 93
50
0
Holidays Plane or Train
Tickets
Day 1‐2 Day 3‐4
12. Case Study – FDM Travels
EKSTRA BLAD
300
DET NETSALG
250
Index Click Through Rate
200
R
155
150 136
no PBT
C
100 100 100 PBT
100 93
50
0
Holidays Plane or Train Household income
Tickets >54.000€ AND
(Holidays OR Plane
or Train Tickets)
Day 1‐2 Day 3‐4 Day 5‐6
13. Case Study – FDM Travels
EKSTRA BLAD
300
DET NETSALG
250
Index Click Through Rate
200
R
155
150 136
no PBT
C
106 PBT
100 100 100 100
100 93
50
0
Holidays Plane or Train Household income Household income
Tickets >54.000€ AND >54.000€
(Holidays OR Plane
or Train Tickets)
Day 1‐2 Day 3‐4 Day 5‐6 Day 7‐8
14. Case Study – FDM Travels
EKSTRA BLAD
300
DET NETSALG
250
Index Click Through Rate
200
183
R
155
150 136
no PBT
C
106 PBT
100 100 100 100 100
100 93
50
0
Holidays Plane or Train Household income Household income Women
Tickets >54.000€ AND >54.000€
(Holidays OR Plane
or Train Tickets)
Day 1‐2 Day 3‐4 Day 5‐6 Day 7‐8 Day 9‐10
15. Case Study – FDM Travels
EKSTRA BLAD
300
280
DET NETSALG
250
Index Click Through Rate
200
183
R
155
150 136
no PBT
C
106 PBT
100 100 100 100 100 100
100 93
50
0
Holidays Plane or Train Household income Household income Women Holiday AND
Tickets >54.000€ AND >54.000€ Household Income
(Holidays OR Plane >54.000€ AND
or Train Tickets) Women
Day 1‐2 Day 3‐4 Day 5‐6 Day 7‐8 Day 9‐10 Day 11‐12
16. FDM Travels: Conversion Rates/Online Sales
EKSTRA BLAD
Not only click through rates increased. Also sales, which was the
aim of the campaign, did.
DET NETSALG
300
275
250
200
Index Conversion Rate
e
no PBT
150
PBT
C
100
100
50
0
17. Case Study – DFDS Seaways
EKSTRA BLAD
Campaign: Rockcruise competition
DET NETSALG
Aim: Click Through Rate
Creatives:
18. Case Study – DFDS Seaways
EKSTRA BLAD
300
DET NETSALG
247
250 239
200
145
150
125 no PBT
118
PBT
100 100 100 100 96 100 100
100
50
0
Top Article Top Article Top Article
Couples 30-49 20-29
19. EKSTRA BLAD
Overall Findings
Predictive Behavioral Targeting achieves much better
results than ordinary campaigns
DET NETSALG
Creatives and target group must be harmonized
Targeting criteria m st be combined smartl
must smartly
Targeting helps the advertiser redefine/optimize the
target group
Publisher can offer the advertisers a better product
with optimized inventory