Transcript Why is Online the New Primetime?

MEASURING THE DIGITAL WORLD
Online is the New Primetime
Gian Fulgoni
Chairman
Web Content 2008
June 18, 2008
Why is Online the New Primetime?
■ It’s where the audience is.
■ There are powerful new tools for online media planning and analysis
■ We have proof of the power of online advertising:
– Correct metrics
– Can measure latent impact
– Can show impact on offline (e.g. in-store) sales
– Can demonstrate long term branding benefit
■ Budgets rapidly shifting online as effectiveness and efficiency of online
marketing is demonstrated, but Internet marketing still underutilized by
many industries
Proprietary and Confidential Do not distribute without written permission from comScore
2
comScore Customer Knowledge Platform:
A 360° View of 2 Million Global Internet Users
Passively Observed
Behavior and Survey Measurement
Designed to be representative of the online population
 Projectable to the total U.S. population
 Ernst & Young certified for information privacy & security
WEB VISITING & VIEWING
DEMOGRAPHICS
•
•
•
•
• Self-reported and validated
• Appended Segments (e.g. Claritas, Acxiom)
• Individual & Household Level
All Web Site/Page Click Stream
Content Viewed
Search Engine Queries
Keyword Used
ONLINE TRANSACTIONS
SURVEYS
• All Secure Session Activity
• Purchases and Subscriptions
• Price Paid, Shipping & Handling,
Promotions
• Applications/Configurations
• E-mail or Contextual “Pops”
• Behavior-activated Surveys
• Observation of All Surveys Taken
Across All Suppliers
OFFLINE PURCHASING
MARKETING STIMULI
• Online Ads
• Referral Links
Proprietary and Confidential Do not distribute without written permission from comScore
• Linked using Name and Address
• Client CRM Databases
• Retailer Loyalty Card Data
• IRI Scanner Panel Data
The Internet is a already a massive medium, but
still growing rapidly in a variety of ways
■ 191 million U.S. people age 2+ online in a month, spending an
average of 30 hours online per person*
– But 77% of 841 million global Internet users now outside of U.S.
■ 99% of online population search in a month,
conducting 80 searches per searcher**
■ 73% of online population stream a video, viewing an average of 83
videos per viewer per month***
– Up 50% vs YA
■ 67% of online population visit a social networking site, spending 4
hours per month per visitor*
■ Online consumer spending totaled $200B in 2007, up 20% vs. YA
– Non-travel represents 8% of consumer spend (excl. food, gas, and autos)
Proprietary and Confidential Do not distribute without written permission from comScore
User Generated Content Sites Growing Rapidly
Monthly Unique Visitors (000)
80,000
70,000
1 [M] MYSPACE.COM*
60,000
2 [M] YOUTUBE.COM
People
(000)
50,000
3 [M] Blogger
40,000
30,000
4 [P] FACEBOOK.COM
20,000
10,000
0
Proprietary and Confidential Do not distribute without written permission from comScore
5
The Internet is the 3rd most used medium and
about to become 2nd only to TV
Internet* 17%
Radio 19%
Video or
DVDs 12%
* Excludes work usage.
Source: Forrester Research, March 2007
Proprietary and Confidential Do not distribute without written permission from comScore
Newspapers
8% Magazines
7%
Watching TV
37%
Internet’s reach exceeds TV’s from 7 am to 8 pm
45
40
% People 15+ Using Medium
Internet
35
30
25
TV
20
Internet
15
TV
10
5
0
Source: National People Meter, comScore Media Metrix 2007
Proprietary and Confidential Do not distribute without written permission from comScore
7
Internet Ad Revenues Continue to Surge
Quarterly Internet Advertising Revenues
($ Billions)
1998
1999
2000
2001
2002
2003
$21.1B in 2007,
Up 26%
Over 2006
2004
2005
2006
2007
Source: Interactive Advertising Bureau (IAB), PricewaterhouseCoopers (PwC) estimates.
Proprietary and Confidential Do not distribute without written permission from comScore
8
Ad Dollars are Flowing Online
Media
Advertising Dollars
Spent in 2007*
Internet
$21 Billion
Outdoor
$7
Cable TV
$26
Magazines
$14
Direct Mail
$61
Miscellaneous
$39
Overall
$295
Trade Journals
$4
Yellow Pages
$14
Broadcast TV
$46
Radio
$19
Newspapers
$44
Percent Change from 2006
Source: Global Insights as published in BusinessWeek
* Other sources may define categories differently, leading to different results
Proprietary and Confidential Do not distribute without written permission from comScore
60% of all ad
spending in
traditional media
goes to branding
campaigns
Yet, online is only 7% of total ad spending!
2007 Estimate
($ in Millions)
% Total
Direct Mail
$
60,988
21%
Broadcast TV
$
45,749
15%
Newspaper
$
42,939
14%
Internet
$
21,100
7%
Cable TV Networks
$
20,479
7%
Radio
$
18,592
6%
Yellow Pages
$
14,538
5%
Consumer Magazine
$
13,695
5%
All other
$
55,977
19%
Total
$ 296,100
100%
Source: Advertising Age by Robert J. Coen, McCann-Erickson Inc; Newspaper Association of America;
Radio Advertising Bureau; Simba Information; Outdoor Advertising Association of America; Interactive
Advertising Bureau (IAB); Lehman Brothers
Proprietary and Confidential Do not distribute without written permission from comScore
10
U.S. Online Advertising:
One third branding; two thirds direct response
•Search is the biggest piece of online advertising
•Branding spending still under-developed
Total Online Advertising 2007: $21.2 Billion
Other 9%
Rich 10%
Display
24%
Classifieds
16%
Search
41%
Display
Search
Classifieds
Other
Rich / Video
Source: Interactive Advertising Bureau (IAB), PricewaterhouseCoopers (PwC) estimates.
Proprietary and Confidential Do not distribute without written permission from comScore
11
The Web Already Delivers More Advertising Than TV!
Monthly GRP’s
Monthly Impressions Per Person
5,000
300.00
250.00
4,000
1.4 x
200.00
2.2 x
3,000
150.00
2,000
100.00
1,000
50.00
-
-
Internet
TV
Internet
TV
The CPM’s are just much, much lower for Internet.
Source: Display Impressions: comScore Ad Metrix, Jan -08, Television GRP’s: comScore Estimates
Proprietary and Confidential Do not distribute without written permission from comScore
12
Frequent news about moves online
■ “Online is getting to the
point where it may be more
important than the 30second TV spot.”
Joel Ewanick
VP Marketing
Hyundai Motor America
■ “The country’s third-largest
advertiser (General Motors) is
getting ready to shift fully half
of its $3 billion budget into
digital and one-to-one
marketing within the next 3
years.”
Advertising Age
March 17, 2008
• Even ad-spending giant P&G is considering cutting its overall ad
budget by as much as 10% this year and aggressively moving ad
dollars to the Web.
WSJ May 12, 2008
Proprietary and Confidential Do not distribute without written permission from comScore
13
New Tools Put Online Media Planning and
Analysis on a Par with other Media
■ Panel-based measures correct problems with site server data
■ Using the correct metrics allows improved media planning and analysis
– Also, look beyond clicks
■ Intelligence is now available showing advertisers and publishers
detailed data on all online ad campaigns
■ Detailed segmentation of users is helping build more powerful media
plans
Proprietary and Confidential Do not distribute without written permission from comScore
14
The Cookie Deletion Problem: Overstated Reach and
Understated Frequency
A comScore Study of Yahoo and DoubleClick Cookies
■ 30% of Internet users delete their cookies in a month
■ These deleters do so an average of 4 times a month
– 5 different cookies for same site in a month on one computer
■ True for 1st party site cookies and 3rd party ad serving cookies
■ Cookie deletion creates large errors:
– 2.5 times overstatement of unique visitors in server logs
– 2.5 times overstatement of reach and a similar understatement of frequency in ad
server logs
■ Need to adjust basic site server statistics using panels
Proprietary and Confidential Do not distribute without written permission from comScore
Looking Beyond your Site Server Data
What’s the Size of my Lost Opportunity?
A: Lost Opportunity Analysis
Analysis of Marriott Visitor Non-Buyers (VNBs)
Dollar Share of Lost Sales Among VNBs
8%
Bought at
Marriott
Other
13%
Had No
Other
Travel
Activity
12.9
Million
Unique
Visitors
92%
Became
Visitor
NonBuyers
(VNBs)
Hilton
19%
Priceline
25%
7%
7%
14%
Hotels.com
87%
That Visit
Another
Travel
Site
88%
Did Not
Convert
IC Hotels
8%
Choice Hotels
10%
Expedia
10%
Starwood
12%
Bought
Elsewhere
$462M Sales Opportunity Gap
Proprietary and Confidential Do not distribute without written permission from comScore
16
Decline in Online Ad Click-Through Rates
2.50%
2.50%
2.00%
1.50%
2002
1.12%
2006
1.00%
0.41%
0.50%
0.20%
0.00%
Rich Media
Sources: Doubleclick, eMarketer, Eyeblaster, ABI Research estimates
Proprietary and Confidential Do not distribute without written permission from comScore
Non-Rich Media
“Natural Born Clickers”
comScore Study with Tacoda and Starcom
Clickers on Display Ads
Source: comScore, Total US Online Population, July 2007
Clickers
Non-Clickers
Proprietary and Confidential Do not distribute without written permission from comScore
Clickers Follow the 80/20 Rule
Source: comScore, Total US Online Population, July 2007
Clickers are predominantly
younger (25 - 44) with lower income (under $40K)
Proprietary and Confidential Do not distribute without written permission from comScore
Those Who Viewed the Ad But Did Not Click Drove
Majority of Site Engagement and Sales
% of Exposed People
Include View-Through Effects When Measuring the
Campaign – Clicks Are NOT Enough
Proprietary and Confidential Do not distribute without written permission from comScore
Majority of Sales Impact of Display Ads
Occurs Offline
Impact of Display Ads
Offline
Online
Online
32%
Offline
68%
Proprietary and Confidential Do not distribute without written permission from comScore
Clicks on Display Ads Are a Misleading Metric and
They Don’t Reflect Brand-Building Effects
■ Only use for direct response ad campaigns (or
search)
■ Clicks don’t measure all of a campaign’s sales
impact, including the cumulative (latent) and offline
impact of ads
■ Clicks don’t tell you anything about brand building
effects
Proprietary and Confidential Do not distribute without written permission from comScore
22
New competitive advertising intelligence
■ comScore Ad Metrix
– Number of ad view impressions…per advertiser…per site
– Precise reach and frequency, full demographics and behavioral
stats
– Transcription of actual ads
– Post-buy analysis
Proprietary and Confidential Do not distribute without written permission from comScore
23
Auto Industry Ad Leaders
Total
Display Ad
Advertiser Views (000)
General
Motors
Toyota
Ford
Motor
Company
Share of
Voice*
AdExposed
Unique
Visitors
(000)
Reach
Average
Frequency
1,687,065
1,356,782
33.00%
26.00%
102,574
62,428
56%
34%
16.4
21.7
1,075,831
21.00%
94,987
52%
11.3
Source: comScore Ad Metrix, Jan 08
Proprietary and Confidential Do not distribute without written permission from comScore
24
Automakers using Portals and Auto Resource
Sites
Top Properties where Auto Manufacturer Display Ads Appear
Ad Exposed
Total
Unique
Display Ad
Visitors
Views (000)
(000)
Reach
Average
Frequency
Yahoo! Sites
935,550
72,041
39%
13.0
Microsoft Sites
584,889
39,615
21%
14.8
Fox Interactive Media
550,288
41,660
23%
13.2
AOL LLC
316,772
33,808
18%
9.4
AutoTrader
113,764
3,698
2%
30.8
Edmunds.com
97,579
4,359
2%
22.4
KBB.com
83,529
4,669
3%
17.9
eBay
80,918
15,027
8%
5.4
Time Warner - Excl. AOL
53,674
8,855
5%
6.1
Google Sites
50,306
16,192
9%
3.1
Source: comScore Ad Metrix, Jan 08
Proprietary and Confidential Do not distribute without written permission from comScore
25
Top Sites Deliver Far Less Than Their Nominal Reach
and Need Ad Networks to Expand Effective Reach
At Least Half of Top Sites Audiences Are Hard to Reach by a Typical Campaign
Reach of
U.S.
Internet
Audience
74.4%
72.4%
58.7%
56.5%
Bottom 50%
of Viewers
for Each Site
Yahoo! Sites
Google Sites
AOL Media Network
MSN-Windows Live
% of Page Views Viewed by Bottom 50% of Site Users
% of Total
Page Views
3.1%
Yahoo!Sites
Sites
Yahoo
Average PVs
per Light User
(bottom 50%)
10.3
3.8%
2.5%
Google Sites
Google
Sites
14.7
MSN-Windows
Live
AOL
Media Network
7.4
2.1%
AOL
Media
Network
AOL
MSN-Windows
Live
7.3
Source: comScore Custom Analysis, December 2007
Proprietary and Confidential Do not distribute without written permission from comScore
26
Why have so many advertisers been slow to
adopt online?
■ Are the wrong metrics being used, i.e. clicks instead of ROI?
■ Is it lack of proof that online advertising works?
■ Is there a bias that online is only for direct response?
■ Do some advertisers not realize what paid search can do for them?
■ comScore’s clients have been busy proving the results
– Search
– Display advertising
– Offline impact (in-store sales)
– Latent effects
Proprietary and Confidential Do not distribute without written permission from comScore
27
“The Role of Search in Consumer Buying”
2006 comScore Study with Google
The Three Components of how Search Drives Buying
Direct Online Effects
16%
Latent Online Effects
21%
Latent Offline Effects
63%
Source: Google/comScore “The Role of Search in Consumer Buying” Press Release, March 21, 2006;
based on holiday-related purchases completed online and offline across 11 product categories for 60
days post-search
Proprietary and Confidential Do not distribute without written permission from comScore
28
“The Digital Shelf”:
comScore Study with P&G, Yahoo and SEMPO
THE SEARCH
MARKETING
OPPORTUNITY
All CPG Categories
% INTERNET
REACH
UNIQUE
VISITORS
UVs USING
SEARCH
Food
93.7MM
43.8MM
Baby
26.0MM
15.7MM
Nearly half of
Internet Users
Personal
search on CPG
Care
related terms or visit
CPG related sites.
35.9MM
9.8MM
7.3MM
1.7MM
Household
USING SEARCH
Source: comScore Marketing Solutions; 3 Months Ending April 2007 – Total U.S.
Proprietary and Confidential Do not distribute without written permission from comScore
29
Searchers are more affluent, younger
and more likely to be female
Demographic Profile
SEARCHERS
U.S.
POPULATION
$63k
$45K
41
46
Female
78%
50%
4 yr. College or Better Education
45%
22%
Average Income
Average Age
Demographic & Life Stage Questions: Base: Searcher (n = 2,011); Non – Searcher (n= 803)
= Significant difference at 95% confidence
Proprietary and Confidential Do not distribute without written permission from comScore
30
Searchers can be influenced to switch brands
based on factors other than price
Thinking about your last purchase, did you switch brands?
Why did you switch?
Price
(Main Reason for Switching)
• Searchers 27%
• Non-Searchers 38%
Searchers
Non-Searchers
QA14/11a: Did you switch to another brand? Base: Searchers (n = 1,893); Non-Searchers (n = 751)
QA15/12a: Why did you switch? Base: Searchers (n = 625); Non-Searchers (n = 208)
= Significant difference at 90% confidence.
Proprietary and Confidential Do not distribute without written permission from comScore
31
Proprietary and Confidential Do not distribute without written permission from comScore
32
The Brand Lift of Search
Methodology: Testing the Brand’s Placement
Subjects entered an unbranded query
and saw one of the following search
engine results pages (SERP):
1. Control (No branding –
“Brand X” does not appear
on the SERP)
2. Brand appears in Top
Sponsored result only
3. Brand appears in Top
Organic result only
1
Control, brand
not mentioned
2
Query: Unbranded (i.e., “fuel efficient
car”)
Top
Sponsored Only
Top
Sponsored
3
Top Organic
5
Side
Sponsored
Top Organic Only
4. Brand appears in both
Top Sponsored and
Top Organic results
5. Brand appears in Side
Sponsored result only
4
Note: All subjects (N=2722) are 25+ and considering
purchasing a new car within the next year (test
brand: major import auto maker).
Top Organic AND
Top Sponsored
Source: Enquiro Search Solutions, Inc., July 2007.
Proprietary and Confidential Do not distribute without written permission from comScore
34
A 16% Point Increase in Brand Association When
Brand Is in Top Sponsored and Top Organic Results
When you think of fuel-efficient cars, which come to mind?
Query: Unbranded (i.e., “fuel efficient cars”)
16%
70%
60%
50%
42% point gap
between test
brand and
another brand
40%
30%
20%
Test Brand
10%
Other
None
0%
Control
Side
Sponsored
Top Organic
Listing
Top Sponsored Top Sponsored
Listing
& Top Organic
Source: Enquiro Search Solutions, Inc., July 2007; Note: There is a significant (7% point) increase from Control to Side Sponsored, as
well.
Proprietary and Confidential Do not distribute without written permission from comScore
35
A 2.2x Lift in Aided Brand Recall When Brand Is in Top
Sponsored and Top Organic Results
Which of the following brands do you remember
seeing in the search results page you just viewed?
Query: Unbranded (i.e., “fuel efficient cars”)
Test Brand
2.2x
70%
1.5x
60%
50%
40%
x
30%
20%
Test Brand
10%
Other
0%
Side
Sponsored
Top Organic
Listing
Top
Top Sponsored
Sponsored & Top Organic
Source: Enquiro Search Solutions, Inc., July 2007.
Proprietary and Confidential Do not distribute without written permission from comScore
36
What about the impact of
Internet ad campaigns on instore sales?
Proprietary and Confidential Do not distribute without written permission from comScore
37
Close the Loop Studies (Done with Yahoo!)
■ Dozens of studies to assess the impact of paid search and banner ads
on online and offline sales
■ Real world analysis: comScore panelists divided into two matched
groups (exposed and non-exposed to advertising)
– Search only
– Display ads only
– Search and display ads together
– Neither
■ Passively measured behavior, no surveys involved
– Linked to in-store sale through CRM databases, retailer loyalty cards (e.g. Kroger),
IRI scanner panel
Proprietary and Confidential Do not distribute without written permission from comScore
38
Search + Display Ad Effectiveness Results:
comScore Norms for Offline Sales Lift
Search ads generate a greater lift in offline sales, but have a far lower reach
and cost substantially more than Display ads
Offline $$ Per (000) Exposed
60%
Control
Incremental Impact
On Offline $’s
Test
50%
40%
30%
% of All
Exposed
Users
20%
8%
11%
80%
10%
0%
Search &
Display
Proprietary and Confidential Do not distribute without written permission from comScore
Search
Display
39
Some Summary Thoughts for the Future
■ The move to online advertising is in full gear
■ It will spread rapidly as traditional advertisers understand its impact
■ Panel data needed for accurate consumer metrics
■ Reach and cumulative frequency are mandatory metrics if traditional
branding ad dollars are to be moved online
■ Measure latency and offline impact of online campaigns to show true
(higher) ROI
■ Search can be a branding tool
Proprietary and Confidential Do not distribute without written permission from comScore
40