Презентация RTB&Programatic Marketing в Украине

1. Gemius Digital Beyond Borders 2015:
How to Equip For The Successful Use of
Programmatic Marketing
Kiev, 06/17/2015

2. Confidential | 2
Agenda
• Who is DataXu?
• Introduction to Programmatic Marketing
• Programmatic Industry
• How it Works
• Programmatic Adoption
• Success Stories
• 5 Factors of Success
• Q&A

3. Who is DataXu?

4. Confidential | 4
Leaders in Industry Innovation
Successful
#1 Fastest Growing
Advertising & Marketing
Company
Scientific
Data Science born
out of MIT and Used
by NASA
Seasoned
Founding team
still at the helm
“DataXu is the number
one offering in market”
Innovative
Operating in 50
countries
Global
Born as a software
company, not an ad
network
Aligned
Lighting the Way to Marketing ROI

5. Confidential | 5
“Strongest Current Offering”
“By trying to solve marketer challenges that go
beyond digital media buying, DataXu has set
itself apart as a more complete solution for
marketers… [DataXu] has the most unique
tool for helping marketers plan offline and
online investments at the most granular level.”
-Forrester Research Inc
The Forrester Wave™: Demand-Side Platforms
(DSPs) Q2 2015
The Forrester Wave™ is copyrighted by Forrester Research, Inc. Forrester and Forrester
Wave™ are trademarks of Forrester Research, Inc. The Forrester Wave™ is a graphical
representation of Forrester's call on a market and is plotted using a detailed spreadsheet with
exposed scores, weightings, and comments. Forrester does not endorse any vendor, product, or
service depicted in the Forrester Wave. Information is based on best available resources.
Opinions reflect judgment at the time and are subject to change.

6. Introduction to Programmatic
Marketing

7. Confidential | 7
 An automated marketing approach
 that uses data, analytics, and real-time software
 to deliver the most relevant message
 to the right person
 in the right context
 for optimal return on investment.
Definition: Programmatic Marketing

8. Confidential | 8
The Digital Lifestyle
launched the age of the
consumer

9. Confidential | 9
Consumer behavior has gone “digital.”

10. Confidential | 10
Juggling connected devices is the norm.

11. Confidential | 11
Changes in Consumer Behavior:
Who is the Omni-channel consumer?
Sports
Sports
Morning Afternoon Evening
41% of device switching comes from users trying to
continue the same task.*
… but marketers lose track of the consumer when they
switch devices
*Source: Facebook Programmatic IO Conference
The average person has 4 devices

12. Confidential | 12
What Marketers Want
Conversion Path
in Real Time
See multi-device
conversion
and path analysis
Improved
Efficiency
Deduplication
of Channels
Link TV
Campaigns
Bridge
digital and TV
campaign data
Message
Control
Sequence
messages across
devices
Accurate
Frequency
Unified counting
across all
devices

13. Confidential | 13
Channel-Centric Planning is Inefficient
Goals: Last-Click CPA
Vendor/Publisher: The Trade
Desk, AdRoll, Criteo, GDN
Goals: Brand Lift, Completed
Views
Vendor/Publisher : TubeMogul,
AdapTV
Goals: CTR
Vendor/Publisher :
Millenial, Fiksu
Siloed Channels make it difficult to switch budgets, optimize, and report
Goals: GRPs
Vendor/Publisher: MSO
100%
TV
10%60%
Display
30%
Video Mobile
TV Budget Digital Budget

14. Confidential | 14
“People-centric” puts faces over places.

15. Confidential | 15
Real-time Customer Engagement Today
Customer-Centric Messages across all formats & devices is essential

16. Confidential | 16
Introduction to Programmatic Marketing
Demystifying Programmatic Marketing

17. Programmatic Industry

18. Confidential | 18
The Ad-Technology Value Chain
Ad Inventory
Sources
Demand Side
Platforms (DSP)
Holding
Company
Agency
Trading Desk
Data Providers
Ad Verification Technologies
Premium
Publishers
Ad Networks
Ad Exchanges
Supply Side
Platforms (SSP)
Agency
AgencyBrand Customer
Advertiser
More:
1) Lumscape (“The Original”)
2) EU Ecosystem

19. Confidential | 19
Ad Networks vs. Exchange vs. DSP
• Typically purchase inventory in blocks, at fixed
prices, not at an impression level
• Arbitrage inventory – buying low from publishers
and selling high to advertisers
• May purchase low-cost inventory on
exchanges and mark up by as much as 40%
• No impression-level decisioning
• No cost transparency
• Open marketplace that services buyers and
sellers in a bidding environment
• Can integrate with a DSP to layer 3rd party data
• Exchanges make up one part of several inventory
sources accessible via DSPs
• Charge a transaction fee for each impression
• Software platform used to access Inventory
Sources, Data Partners, and allow for
optimization, bidding, and reporting in real-time,
hundreds of thousands of times per second
• Decision algorithms allow campaigns to "learn"
and optimize
• Charge software licensing fees, either based on
usage or percent of spend
• Can provide completely transparent costs
and impression-level decisioning
Ad Exchange
• “Yield Manages” unsold direct inventory at lower
CPMs than direct
• Can send impression to an Ad Network
• Can send the impression to the “Exchange” for
bidding or private deals through real time bidding
(RTB)
• Optimization algorithms are applied to maximize
publisher benefit
• Charge a transaction fee for each impression
Ad Network
Sell Side Platform (SSP) Demand Side Platform (DSP)

20. How it Works

21. Confidential | 21
The Programmatic Difference
Traditional Execution
• Plan and revise quarterly
• Buy blocks of impressions
• Fixed price dynamics
• “Averaged” audience insights
• Static
• Faith & Opacity
Programmatic Execution
• Execute and optimize in real time
• Buy individual impressions
• Dynamic pricing efficiencies
• 1:1 consumer insights
• Dynamic
• Science and Transparency
$
$$
$$$
$$$

22. Advertisers Buyers Sellers Publishers
Consumer attention is a biddable asset.

23. Confidential | 23
Real Time Bidding: How it Works
Denys & Yulia
go to olx.ua on
their devices
Request
ad from
publisher
ad server
Redirect
to ad
exchange
Send to
real-time
bidders
• DataXu
• Bidder 1
• Bidder 2
Winner
(DataXu)
serves ad
Each
person sees
a unique ad
Yulia sees a L’Oréal ad
Denys sees a Kia adDenys
Yulia

24. Confidential | 24
Classic planning is based on fixed values
- Volume-time-booking
Budget Impressions Clicks
Real-Time-Advertsing is dynamic and only has the budget as fixed value
- Budget-(time)-booking
Conclusion: Real-Time-Advertising optimizes towards maximum ROI.
Important: Additional adserving costs (due to overdelivery) need to be taken into consideration.
ToDo: Mediaplans need to adapt to dynamic models.
dCPM/IMPS eCPC/Clicks
Budget
Traditional Planning vs. Programmatic

25. Confidential | 25
Runtime in days
CPM in €
60% 50%
Network‘s effective CPM
Client‘s fix CPM Client‘s max. CPM
Effective CPM incl. DSP fee
Buying advantage
Runtime in days
CPM in €
Traditional Planning Programmatic Planning
Traditional Planning vs. Programmatic

26. Confidential | 26
How Does The DSP Select The Ad?
Determine a brand’s unique attributes and
appeal in real-time, then optimizes a precise
match for every impression
Standard & Custom Variables
1. Excel in thin data environments to
maximize reach of decisioning models
2. Use champion-challenger system to
continuously improve performance
3. Out-of-the-box algorithm variations tuned
for specific use cases
• Concept
• Unit size
• Media type
• …
C R E AT I V E
C O N S U M E R
• Location
• Interaction with
advertiser’s site & ads
• Demographics
• Purchase intent
behavior
C O N T E X T
• Content category
• Publisher
• Exchange
• Time of day
• Day of week
• …

27. Why would one buy media this way?
Total price for bag of apples: $100? Price per red apple: $5 Price per green apple: $1

28. Confidential | 28
• The digital media ecosystem has evolved to
enable more granular measurement
• Brands have learned that all impressions are
not created equal, even on the same website
• Paying a flat rate for large blocks of media
enables arbitrage and sell-side profit margin
• Buyers capture significant gains by evaluating
and purchasing impressions one at a time
• Transparent, impression-level decisions drive
efficiency and effectiveness for the brand
Why do brands buy media this way?

29. Confidential | 29
Why would one sell media this way?

30. Confidential | 30
Digital media is highly dynamic.

31. Confidential | 31
• Digital media supply has rapidly outgrown
demand
• Fragmentation of consumer attention has
eroded direct sales leverage
• Digital media supply is dynamic and perishable
• Publishers are increasingly embracing
programmatic, and not just long-tail websites
• Innovative transaction models have emerged to
offer “premium programmatic” options
Why do publishers sell media this way?

32. Confidential | 32
Programmatic: From Direct to Exchange
Private
Exchanges
Direct w/
Programmatic
Exchange
Media
Direct Media
Guaranteed Direct
Publisher Contracts
Exchange cleared
with special
prioritization
Open Exchange
Media Buying
Guaranteed Direct
Publisher Contracts
What are they?
Who can
participate?
How is it
accessed?
How is it
priced?
Who do you
pay?
Advertisers with
direct contracts
Advertisers with
direct contracts
Advertisers approved
by publisher
Advertisers approved
by exchange
Via ad server tags Via ad server tags Via an exchange Via an exchange
What inventory
is typical?
Direct publisher
inventory
Direct publisher
inventory
Prioritized publisher
inventory
Exchange inventory
Fixed price.
Determined by IO
Fixed price.
Determined by IO
Auction with floor or
Fixed Price. First
look option
Auction priced
Advertiser pays
publisher for media
Advertiser pays
publisher for media
and DSP tech fees
Advertiser pays
DSP for media and
any tech fees
Advertiser pays
DSP for media and
any tech fees

33. Confidential | 33
Complete Media Buying Everywhere
Programmatic buying across 100% of your digital media plan.
• Chosen environment
• Chosen placements
• Audience Targeting
• Right ad
• Pre-negotiated price
• Chosen environment
• Audience targeting
• Auction Price with Minimum
• No guarantees
• Contextually relevant
• Audience Targeting
• Auction Price
• Display, mobile, social, &
video













Private
Exchange
Open
Exchange
Direct
Buys
 Display, mobile, & video

34. Confidential | 34
Example of Inventory Partners
Display MobileVideo
Latest partner list: Who We Partner With

35. Programmatic Adoption

36. Confidential | 36
Global Programmatic Category Spend
9 11,5
14,5
19 21
25,55
9,5
13,5
18
23
27,5
2013 2014 2015 2016 2017 2018
$21
$28
$36
$44
$53
Spend(Billions)
$14
US
Non-US
“The total programmatic market will grow from $13.6bn in 2013
to $20.8bn in 2014 and to $53.3bn by 2018 on a global basis.”
Magna Global: The Evolution of Programmatic Buying, September 2014

37. Confidential | 37
Programmatic Status Quo: Global
Source: May 2014, Magna Global

38. Confidential | 38
Programmatic is Scaling Rapidly: Europe
Source: May 2014, Magna Global

39. Confidential | 39
1) Programmatic active in EU5 and Nordics for several years, still new in
CEE and Southern Europe
2) Agency Trading Desks start with regional hubs
3) Google Display Network and non-exchange Facebook are strongest
players
4) 1st party data critical as less 3rd party data available
5) Local 3rd party data providers expected
6) Mobile and video follow display growth quicker than in other markets as
technologies are developed
Programmatic Growth Considerations:
More: Gemius DİJİTAL HARİTALAR

40. Confidential | 40
Customer Lifetime Value
Mobile Payments and Beacons
Perpetually Persistent IDs
Increasing Marketing ROI
Customized Enterprise Deployments
Programmatic Purchase of All Media
Programmatic TV
Programmatic Trends to Watch

41. Success Stories

42. Confidential | 42
Targeting Tactics Work Together To Fuel Performance
Prospecting
Retargeting
New Leads Driven By DataXu Hand Raisers
Coming to Site
51%46%
3%
Major Auto Brand seeking to retarget
consumers and acquire new leads
3% of users exposed to prospecting &
converted to leads right away
46% of users came to site through
prospecting & were converted to leads
over time by retargeting
51% of users came to site without
prospecting & were converted by
retargeting
Prospecting Plays a Key Role in Driving Conversions
Conversions

43. Confidential | 43
Airline Client Case Study
Viewability Helps Travel Take Flight
Decrease Cost Per Booking (Purchase)
Optimize to viewable impressions in order to drive
true performance and engagement, while
maintaining reach and scale
Viewability* & Performance
47%Increase in
Viewability
36%Decrease in Cost
Per Booking
*Viewability measured by DoubleVerify
Challenge:
DataXu Solution:

44. Confidential | 44
How Retail Clients Benefit From
Programmatic Markting
&

45. Confidential | 45
Campaign Overview
Briefing:
Prospecting
 Increase online sales
Strategy:
Run of Network (RON)
Retargeting
Goal:
Reduce Cost-per-Order
„For many years we‘ve relied on DataXu’s technology and service to achieve
positive results for our clients. Especially with this campaign for a leading retailer,
DataXu exceeded both our expectations and those of the client!”
Stefan Doll, Head of Media, metapeople/metaapes GmbH

46. Confidential | 46
Advanced Insights
On what day and at what time was
the creative performance the best?
What influence does the creative
format have in different regions?

47. Confidential | 47
Successful results
 Over several months, DataXu’s leading RTA-Technology and the experienced German
service team made this campaign a great success  DataXu achieved a CPO that was 16%
under the client’s target CPO.
 With the campaign, this leading German retailer saw a 26% increase in new customers.
 The DataXu platform revealed deep insights into multiple campaign variables, e.g. what
creative performed best in what region or at what time.
In this case, the 468 x 60 ad size in the Nielsen regions 3a and 3b achieved the best
performance.

48. Confidential | 48
How Entertainment Clients Benefit From
Programmatic Marketing
&

49. Confidential | 49
The Brief:
 Amplify reach across
French internet users
 Generate high levels
of engagement while
maintaining a low cost
per view
The Strategy:
Run of Network (RON)
Run on Sites (ROS) –
targeting audiences on
over 30 leading French
video publishers
The Goal:
Drive a high video
completion rate
„We‘ve been closely collaborating with DataXu over the past months. We‘re impressed by the
mix of performance and service DataXu offers, combined with a visionary approach regarding
omni-channel marketing. This video campaign was a great example of how DataXu helped us
to exceed our clients‘ expectations.“ –
Julien Duval, Associate Director, Labelium Inside
Campaign Overview

50. Confidential | 50
Do some creatives work better in
certain regions?
How does the performance vary
during a weekday or during the
weekend?
Advanced Insights

51. Confidential | 51
 DataXu delievered more than 3 million Video Impressions
during the 4 week period & over 60% of users that started
the ad also completed it
92.63%
84.92%
79.03%
60.37%
Watched 1/4 of
Video Ad
Watched 1/2 of
Video Ad
Watched 3/4
ofVideo Ad
Completed Video
Ad
The Results
 DataXu achieved a cost per view (CPV) of 0.01 €, which equals more than 220,000
video completions
 Campaign insights can be as valuable as market research:
- High performance amongst 25-34 year olds
- Best results in the South–East of France during the evening hours on weekends

52. Confidential | 52
950,000
3,068,684
Competitor DataXu
DataXu achieved
an additional
2 million
Impressions
Benefits for the Client
 DataXu was able to reduce the client‘s CPM rate by a third compared to
other providers:
 Achieved more than 60,000 € budget savings or an additional 2 million Ad
Impressions for the same media budget

53. Confidential | 53
Pixum Builds Brand Awareness With
Digital Video
&

54. Confidential | 54
Campaign Goals
 Increase branded searches (rather than generic keyword
searches)
 Lower overall cost-per-conversion
 Generate new customers
Hypothesis
Digital video advertising will build Pixum brand awareness
The Brief

55. Confidential | 55
The Methodology: Pulsing
 Conduct comprehensive effectiveness
analysis
 Campaign modelling based on
generated data
 Run campaign across different
geographic regions
 Employ different pulse (investment)
schedules

56. Confidential | 56
 Test the impact of video advertising on
branded search and website traffic using the
DataXu pulse method
 Measure the number of page views,
searches for “Pixum”, actual software
downloads and sales
 Analyse and compare regional differences
The Methodology: Pulsing

57. Confidential | 57
 Region 1 and 2 had similar outcomes, showing an overall growth in clicks
 The video campaign led to increased search volumes, resulting in a
positive ROI from digital video advertising for Pixum
Region 1 Region 2 Region 3
The Results

58. Confidential | 58
 Total clicks increased as video advertising investment grew
 The model shows that CPC is expected to decrease over time
-10%
-8%
-6%
-4%
-2%
0%
0 50 100 150
0%
10%
20%
30%
40%
50%
0 50 100 150
Video investment (Euro/1M people in a day)
Organic
Paid
Click Volume Growth
In Region 1, organic clicks grew faster
than paid

59. Confidential | 59
 Determined optimal marketing investment
for all three test regions
 Gained new customers
 Detailed campaign analysis
Campaign Results

60. Programmatic Marketing in
Practice: 5 Factors of Success

61. Confidential | 61
1) Campaign Goal And Attribution Model
Source: Out with the Old
Take Aways:
1. Plan, optimize and measure
campaign success based on
actions directly tied to actions
2. Identify actions that
demonstrate significant interest
in your campaign
3. Value the View: Focus on
audiences that have a
propensity to convert

62. Confidential | 62
2) Data Strategy
1. Leverage 1st party data. It’s the advertiser’s data, not the media owner’s.
2. Data strategy example:
1. Learning Pixel:
1. Homepage
2. Landingpage
3. Shopping Cart/ Order Process
2. Conversion Pixel:
1. Soft Conversion
2. Final thank you page
3. Ask for learnings: How does your data influence your marketing investments?

63. Confidential | 63
CONVERSIONAWARENESS CONSIDERATION| |
Prospecting
Retargeting
OMNI-CHANNELTARGET
AUDIENCE
Geofencing
3rd Party Audience Targeting
Contextual & Site Targeting
3) Enhance Performance At All Points
During the Sales Funnel

64. Confidential | 64
Geo
Targeting Retargeting
Device
Targeting
Prospecting/
Look-a-like
Modeling
Audience
Buying
Contextual
Targeting
4) Optimization: Algorithm Marketplace
Algorithm Variations tuned
for specific use cases
Model & Predict behaviors to
reach more customers
Test and Learn to continuously
improve performance

65. Confidential | 65
5) Brand Safety
BRAND SAFETY
LEVEL 4
$$$$
CLIENT
LIST
CLIENT SPECIFIED SITE LIST
LEVEL 3 DYNAMIC
SEMANTIC
WHITELIST$$$
DOES NOT ALLOW
UNCATEGORIZED,
ANONYMOUS, OR IAB
“NON-STANDARD”
INVENTORY.
LEVEL 2
DYNAMIC
SEMANTIC
BLACKLIST
$$
REMOVES CONTENT
IN IAB
“NON-STANDARD”
CATEGORIES.
ALLOWS
ANONYMOUS
INVENTORY.
LEVEL 1
$
AD EXCHANGE FILTERS
BASIC BRAND
SAFETY
PROVIDED BY
EXCHANGES
BLOCKS MOST
OBJECTIONABLE
CONTENT
INCREASING BRAND
SAFETY
DECREASES REACH
AND INCREASES COST
OF MEDIA
Suspicious
Clicks
Quality
Scoring
Malware and
Virus Scans

66. Confidential | 66
• Guaranteed 97% fraud-free “real people” investment
• If you experience a fraud rate greater than 3%, DataXu will:
– Refund 100% of the DataXu fees associated with the fraudulent media >3%
– Refund 100% of all amounts recovered from the suppliers of that media
5) Ad Fraud Protection
World-Wide Omni-
Channel
FreeNo action
required

67. ДЯКУЮJost Löhnenbach
Head of Sales, New and Emerging Markets - Europe
jloehnenbach@dataxu.com
Link to more resources.
Question & Answers