A HISTORY OF PROGRAMMATIC MEDIA
I. Programmatic Evolution
II. How Programmatic Works
Table of Contents
2	
  
PR OGRAMMATIC E VO LU TIO N
RTB GOES
MAINSTREAM
1994 1996 2005 2007 20112000
DIGITAL AD EVOLUTION
DSPs
LAUNCH
AD
EXCHANGES
LAUNCH
ADWORDS
LAUNCH
(SEARCH)
AD
NETWORKS
LAUNCH
DIGITAL
ADS
LAUNCH
4	
  
2008
MDC
LAUNCHES
VMM
Advertiser and Publisher negotiate CPM and budget to show
an ad for a given period of time
Advertiser
Publisher
5	
  
IN THE BEGINNING…
DIGITAL AD BUYING WAS RELATIVELY SIMPLE
AdvertiserPublishers
More parties to deal with / Higher cost to attain reach / Higher overhead
	
  
6	
  
BUT IT BECAME EXPONENTIALLY COMPLEX
Publishers
MARKET OPPORTUNITIES DEVELOPED
•  Inventory centralization
•  Reduced overhead
•  Scale at lower costs
•  Automation
•  Smarter targeting
•  More data
7	
  
Pros:
Advertisers: Centralized
buy; lower cost
Publishers: Helped sell
inventory at scale
Networks aggregate site inventory into single package
Cons:
Advertisers: less transparency and
precision with pubs and audience
Publishers: Still a large amount of
unsold inventory
Tremor
Network
Reuters
Men’s Fitness
Food Network
Sports Illustrated
AOL Network
VH1
USA Today
Ellen
Comedy Central
8	
  
NETWORKS SIMPLIFIED THE PROCESS
A marketplace to make real-time bids for ad inventory (like search)
Pros:
Advertiser: Greater site
flexibility and lower costs
Publisher: Sell larger amounts
of ad space
Cons:
Advertiser: Need technology to manage
multiple exchanges and hit audience
Publisher: Difficult to maximize
profitability
Google (ADX)
Exchange
Yahoo
Exchange
NetworksNetworks
9	
  
EXCHANGES ADDED SCALE
Publishers Publishers
-Focuses on the audience rather than the site
-Centralizes point of access for inventory
-Access to robust data for smarter decisions
Programmatic
Exchange
Network
Publishers
10	
  
PROGRAMMATIC EMPOWERED WITH CONTROL
Programmatic uses intelligent software to purchase digital advertising
PROGRAMMATIC FLIPPING THE MARKET
11	
  
0%	
  
20%	
  
40%	
  
60%	
  
80%	
  
100%	
  
2011	
   2012	
   2013	
   2014	
   2015	
   2016	
   2017	
  
Programma3c	
   Non-­‐Programma3c	
  
eMarketer: “US Display Ad Spending Share, by Type” 2013
HO W
PROGR AMMATIC
WORKS
Definition
(IAB)
Publisher
Advertiser
Programmatic
Guaranteed
(Premium, Direct)
Private Exchange
(Invite-Only
Auction)
RTB
(Open Auction)
TYPES OF PROGRAMMATIC BUYING
13	
  
Purchase of ads through
real-time auctions
Publisher restricts auction to
select buyers/advertisers
Resembles traditional direct
sales; differentiated by
automation of RFP, IO,
trafficking for publisher
Definition
(IAB)
Publisher
Advertiser
Programmatic
Guaranteed
(Premium, Direct)
Private Exchange
(Invite-Only
Auction)
RTB
(Open Auction)
TYPES OF PROGRAMMATIC BUYING
14	
  
Purchase of ads through
real-time auctions
Publisher restricts auction to
select buyers/advertisers
Resembles traditional direct
sales; differentiated by
automation of RFP, IO,
trafficking for publisher
PROGRAMMATIC	
  
RTB	
  
PROGRAMMATIC RTB IN LUMASCAPE
15	
  
•  Traders (trading desks): the people
•  Demand Side Platform (DSP): trading desk tool
•  Data Management Platform (DMP): data
•  Supply Side Platform: publisher tool
COMPONENTS OF PROGRAMMATIC RTB
16	
  
2. Publisher
reaches out to SSP
to sell inventory
DMP	
  
3. SSP sends
bid request with
audience data
(DMP) to DSPsPublisher
Traders Advertisers
$2.00 CPM
$1.50 CPM
5. SSP chooses
winner; advertiser
pays $0.01 more
than competitor
$1.00 CPM
4. Traders send bids to
show ad for audience
(DMP) to DSP
Winner
$1.51 CPM
Programmatic
Programmatic
SSP	
  
1. User lands on
publisher site
PROGRAMMATIC RTB FLOW
DMP	
  
DSP	
  
DMP	
  
DSP	
  
DMP	
  
DSP	
  
•  The people (RTB buyers)
•  An organization that manages programmatic
media typically using DSPs and other audience
buying technologies
–  Place ads based on target audience
–  Manage bids, placements, and targets
TRADERS
18	
  
•  Trading initially very complicated, requiring
specialized expertise from traders
•  Technology has evolved, making
programmatic buying more intuitive
•  New technology empowers agencies and
brands to take back control of
programmatic buys with in-house traders or
search teams
EVOLUTION OF TRADERS
19	
  
•  Tool for trading desk (advertiser)
•  A piece of software used by traders (advertisers) to buy advertising
placements with real-time bids
•  Benefit of DSP for Advertisers
–  Central access to publishers, networks, exchanges, SSPs
–  Targeting/control options
–  Robust reporting
–  Automated optimizers (targeting/ads)
–  Brand protection (white list, content, viewability)
DEMAND SIDE PLATFORM (DSP)
20	
  
Targeting Control
Audience
•  Demographic
•  Interests/Behavior
Ø  First Party
Ø  Third Party
•  Geography
Site
•  List
•  Category
•  Content
•  Budget
•  Bids
•  Day Part
•  Frequency
•  Brand Protection
21	
  
DSP LEVERS
•  Data for both advertisers and publishers
•  A data warehouse that houses information and spits
it out in a way that’s useful for marketers and
publishers
•  Benefit of DMP for Advertisers
–  Data to develop and target specific targets
–  Target specifications applied across large source
DATA MANAGEMENT PLATFORM (DMP)
22	
  
•  Tool for publishers
•  A piece of software used by publishers to sell advertising in
an automated fashion
•  Benefits of SSP for Publisher
–  Access to advertisers, networks, ad exchanges, DSPs
–  Automated yield management
–  Robust reporting
–  Block advertisers
–  Price floors
SUPPLY SIDE PLATFORM (SSP)
23	
  
•  Companies cross lines
–  Rubicon (SSP + Exchange): Advertising Automation Cloud
–  AppNexus (DSP + SSP): Flexible Technology
–  Google (DSP + Exchange + SSP): Internet Monopoly
•  Issues
–  Market confusion
–  Who is more important? Advertiser or Publisher?
•  Bottom Line for Advertisers: Does it work?
24	
  
BLURRED LINES
TACTI CS AND
STRATEG IES
Prospecting
•  Targeting ads using audience or site data
•  Scale new customers
•  Learn about audience
Retargeting
•  Targeting ads based on users’ past site visitation
•  Convert intent into sale
•  Overlay prospecting strategies
26	
  
TACTICS
Control Parameters (Campaign)
Budget: $1,000/day
Max CPM: $10
Frequency: 4/day
Brand Safety: No Sexual Content
Targeting Requirements
Geo: NY/NJ
Day Part: Weekends
Demographics < User Age < 36-45
Demographics < HHI $150k-250k
•  Jewelry brand
•  Goal: Drive qualified traffic and revenue
27	
  
PROSPECTING EXAMPLE
28	
  
PROSPECTING TARGETS
Demographic
s
Social Media Users
Very Highly Engaged
Social Media Users
Very Highly Engaged
Social Media Users
Very Highly Engaged
Interest
Fashion and Style Attitudes:
Timeless fashion
Fashion and Style Attitudes:
Timeless fashion
Fashion and Style Attitudes:
Timeless fashion
In-Market
Women’s Fashion and Apparel
Fine Jewelry
Apparel & Accessories
Luxury Retailers
Clothing & Accessories
Luxury Accessories
Target 1 Target 2 Target 3
Target 1:
Fine Jewelry
Target 2:
Luxury Retailers
Target 3:
Luxury Accessories
Low CPA
Low Cost
High Volume
High CPA
High Cost
High Volume
Low CPA
Low Cost
Low Volume
Metrics Optimization
Increase daily budget
Pause
Increase CPM or frequency
Test without ‘Timeless Fashion’
29	
  
PROSPECTING OPTIMIZATION
NEED PRODUCT DATA 30	
  
RETARGETING STRATEGIES
iPad Page
(no purchase)
Retarget with
iPad Ads
General
Sequential
Messaging
Last Page Visited Search
Apple Page
(no purchase)
Retarget with
iPad Ads
Second Ad:
iPad Sizes (DR)
First Ad:
Lifestyle Ad (TV)
Upsell
Search for
‘tablet’
Retarget with
iPad Ads
Bought
iPad
Retarget with
iPad Cases
PROGR AMMATIC
VIS ION
The majority of media will be bought programmatically through self-
service platforms, which makes media buying more transparent,
flexible, and efficient. But this technology requires a new type of
media professional. With insight into this evolution, The Media Kitchen
has developed the Programmatic Media Team.
The Programmatic Media Team will oversee 3 major areas:
•  Programmatic RTB (includes banners, video, and mobile)
•  Programmatic Direct (includes banners, video, and mobile)
•  Paid Search (e.g. Google and Bing)
•  Paid Social Media (e.g. Facebook, Twitter, LinkedIn)
32	
  
PROGRAMMATIC MEDIA TEAM
THA NK
YOU

A History of Programmatic Media

  • 1.
    A HISTORY OFPROGRAMMATIC MEDIA
  • 2.
    I. Programmatic Evolution II.How Programmatic Works Table of Contents 2  
  • 3.
    PR OGRAMMATIC EVO LU TIO N
  • 4.
    RTB GOES MAINSTREAM 1994 19962005 2007 20112000 DIGITAL AD EVOLUTION DSPs LAUNCH AD EXCHANGES LAUNCH ADWORDS LAUNCH (SEARCH) AD NETWORKS LAUNCH DIGITAL ADS LAUNCH 4   2008 MDC LAUNCHES VMM
  • 5.
    Advertiser and Publishernegotiate CPM and budget to show an ad for a given period of time Advertiser Publisher 5   IN THE BEGINNING… DIGITAL AD BUYING WAS RELATIVELY SIMPLE
  • 6.
    AdvertiserPublishers More parties todeal with / Higher cost to attain reach / Higher overhead   6   BUT IT BECAME EXPONENTIALLY COMPLEX Publishers
  • 7.
    MARKET OPPORTUNITIES DEVELOPED • Inventory centralization •  Reduced overhead •  Scale at lower costs •  Automation •  Smarter targeting •  More data 7  
  • 8.
    Pros: Advertisers: Centralized buy; lowercost Publishers: Helped sell inventory at scale Networks aggregate site inventory into single package Cons: Advertisers: less transparency and precision with pubs and audience Publishers: Still a large amount of unsold inventory Tremor Network Reuters Men’s Fitness Food Network Sports Illustrated AOL Network VH1 USA Today Ellen Comedy Central 8   NETWORKS SIMPLIFIED THE PROCESS
  • 9.
    A marketplace tomake real-time bids for ad inventory (like search) Pros: Advertiser: Greater site flexibility and lower costs Publisher: Sell larger amounts of ad space Cons: Advertiser: Need technology to manage multiple exchanges and hit audience Publisher: Difficult to maximize profitability Google (ADX) Exchange Yahoo Exchange NetworksNetworks 9   EXCHANGES ADDED SCALE Publishers Publishers
  • 10.
    -Focuses on theaudience rather than the site -Centralizes point of access for inventory -Access to robust data for smarter decisions Programmatic Exchange Network Publishers 10   PROGRAMMATIC EMPOWERED WITH CONTROL Programmatic uses intelligent software to purchase digital advertising
  • 11.
    PROGRAMMATIC FLIPPING THEMARKET 11   0%   20%   40%   60%   80%   100%   2011   2012   2013   2014   2015   2016   2017   Programma3c   Non-­‐Programma3c   eMarketer: “US Display Ad Spending Share, by Type” 2013
  • 12.
  • 13.
    Definition (IAB) Publisher Advertiser Programmatic Guaranteed (Premium, Direct) Private Exchange (Invite-Only Auction) RTB (OpenAuction) TYPES OF PROGRAMMATIC BUYING 13   Purchase of ads through real-time auctions Publisher restricts auction to select buyers/advertisers Resembles traditional direct sales; differentiated by automation of RFP, IO, trafficking for publisher
  • 14.
    Definition (IAB) Publisher Advertiser Programmatic Guaranteed (Premium, Direct) Private Exchange (Invite-Only Auction) RTB (OpenAuction) TYPES OF PROGRAMMATIC BUYING 14   Purchase of ads through real-time auctions Publisher restricts auction to select buyers/advertisers Resembles traditional direct sales; differentiated by automation of RFP, IO, trafficking for publisher
  • 15.
  • 16.
    •  Traders (tradingdesks): the people •  Demand Side Platform (DSP): trading desk tool •  Data Management Platform (DMP): data •  Supply Side Platform: publisher tool COMPONENTS OF PROGRAMMATIC RTB 16  
  • 17.
    2. Publisher reaches outto SSP to sell inventory DMP   3. SSP sends bid request with audience data (DMP) to DSPsPublisher Traders Advertisers $2.00 CPM $1.50 CPM 5. SSP chooses winner; advertiser pays $0.01 more than competitor $1.00 CPM 4. Traders send bids to show ad for audience (DMP) to DSP Winner $1.51 CPM Programmatic Programmatic SSP   1. User lands on publisher site PROGRAMMATIC RTB FLOW DMP   DSP   DMP   DSP   DMP   DSP  
  • 18.
    •  The people(RTB buyers) •  An organization that manages programmatic media typically using DSPs and other audience buying technologies –  Place ads based on target audience –  Manage bids, placements, and targets TRADERS 18  
  • 19.
    •  Trading initiallyvery complicated, requiring specialized expertise from traders •  Technology has evolved, making programmatic buying more intuitive •  New technology empowers agencies and brands to take back control of programmatic buys with in-house traders or search teams EVOLUTION OF TRADERS 19  
  • 20.
    •  Tool fortrading desk (advertiser) •  A piece of software used by traders (advertisers) to buy advertising placements with real-time bids •  Benefit of DSP for Advertisers –  Central access to publishers, networks, exchanges, SSPs –  Targeting/control options –  Robust reporting –  Automated optimizers (targeting/ads) –  Brand protection (white list, content, viewability) DEMAND SIDE PLATFORM (DSP) 20  
  • 21.
    Targeting Control Audience •  Demographic • Interests/Behavior Ø  First Party Ø  Third Party •  Geography Site •  List •  Category •  Content •  Budget •  Bids •  Day Part •  Frequency •  Brand Protection 21   DSP LEVERS
  • 22.
    •  Data forboth advertisers and publishers •  A data warehouse that houses information and spits it out in a way that’s useful for marketers and publishers •  Benefit of DMP for Advertisers –  Data to develop and target specific targets –  Target specifications applied across large source DATA MANAGEMENT PLATFORM (DMP) 22  
  • 23.
    •  Tool forpublishers •  A piece of software used by publishers to sell advertising in an automated fashion •  Benefits of SSP for Publisher –  Access to advertisers, networks, ad exchanges, DSPs –  Automated yield management –  Robust reporting –  Block advertisers –  Price floors SUPPLY SIDE PLATFORM (SSP) 23  
  • 24.
    •  Companies crosslines –  Rubicon (SSP + Exchange): Advertising Automation Cloud –  AppNexus (DSP + SSP): Flexible Technology –  Google (DSP + Exchange + SSP): Internet Monopoly •  Issues –  Market confusion –  Who is more important? Advertiser or Publisher? •  Bottom Line for Advertisers: Does it work? 24   BLURRED LINES
  • 25.
  • 26.
    Prospecting •  Targeting adsusing audience or site data •  Scale new customers •  Learn about audience Retargeting •  Targeting ads based on users’ past site visitation •  Convert intent into sale •  Overlay prospecting strategies 26   TACTICS
  • 27.
    Control Parameters (Campaign) Budget:$1,000/day Max CPM: $10 Frequency: 4/day Brand Safety: No Sexual Content Targeting Requirements Geo: NY/NJ Day Part: Weekends Demographics < User Age < 36-45 Demographics < HHI $150k-250k •  Jewelry brand •  Goal: Drive qualified traffic and revenue 27   PROSPECTING EXAMPLE
  • 28.
    28   PROSPECTING TARGETS Demographic s SocialMedia Users Very Highly Engaged Social Media Users Very Highly Engaged Social Media Users Very Highly Engaged Interest Fashion and Style Attitudes: Timeless fashion Fashion and Style Attitudes: Timeless fashion Fashion and Style Attitudes: Timeless fashion In-Market Women’s Fashion and Apparel Fine Jewelry Apparel & Accessories Luxury Retailers Clothing & Accessories Luxury Accessories Target 1 Target 2 Target 3
  • 29.
    Target 1: Fine Jewelry Target2: Luxury Retailers Target 3: Luxury Accessories Low CPA Low Cost High Volume High CPA High Cost High Volume Low CPA Low Cost Low Volume Metrics Optimization Increase daily budget Pause Increase CPM or frequency Test without ‘Timeless Fashion’ 29   PROSPECTING OPTIMIZATION
  • 30.
    NEED PRODUCT DATA30   RETARGETING STRATEGIES iPad Page (no purchase) Retarget with iPad Ads General Sequential Messaging Last Page Visited Search Apple Page (no purchase) Retarget with iPad Ads Second Ad: iPad Sizes (DR) First Ad: Lifestyle Ad (TV) Upsell Search for ‘tablet’ Retarget with iPad Ads Bought iPad Retarget with iPad Cases
  • 31.
  • 32.
    The majority ofmedia will be bought programmatically through self- service platforms, which makes media buying more transparent, flexible, and efficient. But this technology requires a new type of media professional. With insight into this evolution, The Media Kitchen has developed the Programmatic Media Team. The Programmatic Media Team will oversee 3 major areas: •  Programmatic RTB (includes banners, video, and mobile) •  Programmatic Direct (includes banners, video, and mobile) •  Paid Search (e.g. Google and Bing) •  Paid Social Media (e.g. Facebook, Twitter, LinkedIn) 32   PROGRAMMATIC MEDIA TEAM
  • 33.