Transcript Slide 1

Impacts and ROI of Internet Local,
Classifieds, and Directory Advertising
Research and Analysis from comScore Networks, Inc.
March 13th , 2006
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Study Objectives
 Measure the effectiveness of classifieds, directory, and local
advertising on the Internet:
 Across a variety of representative industries
 Incorporating all aspects of ROI including impressions linked to offline
value and branding
 Considering metrics related to sales, awareness, and branding
 Using a high quality methodology that:
 Reflects real market conditions
 Provides a non-artificial test environment to the extent possible
 Offers adequate sample for statistical significance
Why Case Studies?

Each publisher has a unique business model
 Cost-per-lead
 Flat listing fees
 Dramatically different industries

“Marketer” has a different meaning across
each publisher
 Cars.com focus is on “dealerships” collectively
 Verizon Superpages.com focus is on small businesses
organized by category
 CareerBuilder.com focus is on employers of “Top 4
Accounting Firm” and “Leading International Bank”
 Yahoo! Local focused on “Ford Rental”

No comparisons to other mediums
 Cost prohibitive
 Difficult to collect
Participating Publishers and Marketers
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary – Case Studies
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Step 1: Study Details By Publisher

10 publisher and marketer combinations tested
Publisher
Cars.com
CareerBuilder.com
Yahoo! Local
Verizon Superpages.com





Marketers
Car Dealerships Around the Country
Big 4 Accounting Firm and a Major International Bank
Ford Rent-a-Car
Hospitality, Real Estate, Retail, Prof. Services, Autos, Retail
Population:
Passive Data Collection:
Survey Field Dates:
POLK Data Match:
Sample Sizes:
Total U.S.
January 2005 through September 2005
June through September 2005
April 2004 to June 2005
~ 200 survey completes per cell on average (ranging from 138 to 256)
~ 1,000 to 60,000 passive ad exposures per cell
~ 165,000 records matched to POLK automotive database
Step 2: Passively Track Exposure

Example of Listing Exposure at
Step 3: Create Test and Control Cells
Survey Test – Control Design For
 comScoreQ2 conducted 355 Internet surveys in the US from August 31, 2005 through
September 19, 2005.
Definition
Test Group
Control Group
Respondents ages 18 and up who
have used CareerBuilder.com to
search for a job in the past 6
months.
Respondents ages 18 and up who
were aware of CareerBuilder.com,
but who used other sites to search
for a job in the past 6 months.
Sample Size
217
138
Step 4: Automotive database match

POLK covers ~ 70% of new and used car sales in the U.S.
WA
AK
VT
MT
OR
ND
CO
IL
KS
PA
OK
NM
WV
VA
KY
NC
TN
SC
AR
MS
TX
OH
IN
MO
HI
AZ
MI
IA
NE
NV
NY
WI
SD
WY
CA
NH
MN
ID
UT
ME
AL
GA
LA
FL
Full Coverage of new and used vehicle sales
Partial Coverage of new car sales. No used car sales information.
MA
RI
CT
NJ
DE
MD
Step 5: Targeted surveys for offline details
SURVEY FLOW FOR ALL STUDIES
Survey
Section
Screener
metrics
- Eligibility
- Age
- Gender
Awareness
& Usage
Perceptions
Demographics
- Brand
awareness
- Likelihood to
re-use
- Frequency of
local search
-Unaided
- Likelihood to
recommend
- Tenure of local
search
- Attribute
ratings
- Education
-Aided
- Ad recall
- Usage
- Action (s)
taken
- Overall
impression
- Impact of
advertiser
- Annual
household
income
- Zip location
Step 6: Passive Data “Funnel” Analysis:
 Funnel to conversion for Ford rent-a-car
Searched
Keywords
yp.yahoo.com
Online
Clicked
Thru to
Rent Car
Saw Ford Rental
Listing & Clicked
Thru to Dealer
Locator Page Clicked
Thru to
Search for
Availability
Offline
Phone or
Visited
Dealer to
Rent Car
Rented
Car
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Monthly Market Penetration and Size


The majority of the Internet population is now using directories
Even unique life events (like car buying or job seeking) are common
Unique Visitors (MM)
150,000
72% Reach
125,000
100,000
29% Reach
75,000
27% Reach
50,000
25,000
0
Directories/
Resources
Source: comScore MediaMetrix; Total US Internet February 2006
Automotive
Careers
Cumulative Reach by Category


Over a 6 month period, the cumulative reach continues to grow
Offering marketers a huge opportunity to touch consumers
83%
87%
89%
90%
92%
74%
32%
45%
41%
53%
48%
58%
53%
65%
61%
62%
56%
29%
5-Aug
5-Sep
5-Oct
Directories / Resources
Source: comScore Marketing Solutions; Total US Internet January 2006
5-Nov
Careers
5-Dec
6-Jan
Automotive
Five Month Retention By Category


Retention remains strong even over long periods of time
Once engaged, it’s clear consumers value these category resources
87%
85%
58%
55%
5-Sep
83%
50%52%
5-Oct
Directories / Resources
Source: comScore Marketing Solutions; Total US Internet January 2006
81%
49%
45%
5-Nov
Careers
81%
46%
39%
46%47%
5-Dec
5-Jan
Automotive
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary – Case Studies
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Case Study: Yahoo! Local and Ford Rental

Business Model
 Pay-per-listing model
 Objective to drive converting consumers to Ford Rental site

Conversion Funnel






Search for Car Rentals
Clicked on Ford Listing
Availability Check
Requested Rental
Converted Online
100%
40%
31%
5%
3%
ROI Calculation: Total revenue over total cost




Projected consumers exposed to a listing
Conversion both online and offline
Calculate total dollars in rentals for Ford
Divided cost of Yahoo! Local Listing
Case Study: Yahoo! Local and Ford Rental
Conversion Results
 ROI calculations must include offline conversions
3.8%
3.2%
Online
Offline
Total Conversion is 7%
Source: comScore Marketing Solutions; Total US Internet October 2005
Nearly 40% of
Conversion
Occurs offline
Case Study: Yahoo! Local and Ford Rental
Conversion Results
What prevents consumers from converting online?
 Availability and selection of rental cars
 This is a potential revenue loss of ~$1.3 Million* per month
Online
* Monthly UV’s exposed to Ford x drop of rate during availability check x average revenue per car rental
Case Study: Yahoo! Local and
Ford Rental Cars ROI Conclusion


The average price of a rental car transaction is $130 according comScore
eCommerce data
Ford ROI is strong even at much lower average transaction sizes
$$ Per Rental
7.8X ROI
8.7X ROI
$117
$130
6.5X ROI
$100
Source: comScore Marketing Solutions; Total US Internet October 2005
Case Study: CareerBuilder.com,
Accounting Firm and International Bank

Business Model
 Pay-per-listing model
 Objective to connect employers with qualified job seekers

Conversion Funnel





Saw Job Listing
Applied online and offline
Interviewed
Received Job Offer
Accounting Firm
100%
35%
13%
2%
Bank
100%
36%
20%
14%
ROI Calculation: Cost Per Job Offer




Job seekers viewing listing pages
Conversion to application, interview, and finally job offer
Total cost of job listings
Costs divided by offers made = Cost Per Job Offer
Case Study: CareerBuilder.com Cost
Per Job Offer Made ROI Calculation


Cost per job offer made based on listing fees of $389
Differences driven by number of listings, job appeal, and brand
equity
$283
Average length of time to fill a
position: 30 Days
Source: The EDGE Report – August 2005
Median permanent placement fee via
staffing firm/recruiter: $8,592
Source: Staffing Analysts Industry, Inc. - 2005
$30
Accounting Firm
Source: comScore Marketing Solutions; Total US Internet October 2005
International Bank
Case Study: Verizon Superpages.com

Business Model
 Pay-per-listing model
 Objective to connect consumers with a local service provider

Conversion Funnel





Viewed listings page
Clicked listings detail page
Took an online action
Took an offline action
Retail Category Example
100%
25%
8%
17%
ROI Calculation: Total revenue over total cost




Projected number of listings viewers
Number of leads converting online and offline
Calculate total revenue generated
Divided total cost of listings
Case Study: Verizon Superpages.com
Site Reach By Category
 Majority of Listings Views are in Hospitality (41.7%)
 Higher consideration categories are less frequented
41.7%
Site Reach By Listings Category:
Based on Search Activity
29.2%
16.6%
6.8%
Hospitality
Professional
Services
Retail/Shopping
Source: comScore CMS, Total Internet Population, Average Monthly Estimates
Base: ~60,000 Directories Users between January 2005 July 2005
Automotive
4.0%
1.8%
Real Estate
Construction
and Contractors
Case Study: Verizon Superpages.com
Total Conversion By Category
 Conversion varies dramatically by category
 High conversion categories linked to an “immediate need”
Total Conversion By Listings Category:
Based on Listing Page Exposure
25.0%
Retail
25.0%
Automotive
21.3%
Hospitality
Source: comScore CMS, Total Internet Population, Average Monthly Estimates
Base: ~60,000 Directories Users between January 2005 July 2005
13.3%
12.7%
Professional
Services
Real Estate
6.4%
Construction &
Contracters
Case Study: Verizon Superpages.com
ROI Summary By Category
 Average ROI is across all categories $10.60 for every $1 spent
$25
$22.54
$18.85
$20
$13.60
$15
$10
$5
$1.93
$1.21
$0.31
$0
Hospitality
Retail
Real Estate
Source: comScore Marketing Solutions; Total US Internet October 2005
Automotive
Prof. Services
Construction &
Contractors
Case Study: Cars.com

Business Model
 Subscription model
 Objective to connect dealers with
prospective car buyers

Conversion Funnel
 Viewed dealer listing pages
 Searched for vehicle
 Viewed detailed vehicle listings
 Contacted a dealer
 Converted Offline

100%
38%*
ROI Calculation: Cost Per Sale
 Number of unique visitors viewing dealer listings
 Percentage converting based on POLK data match
 Number of vehicle sales to Mid/Low funnel visitors
 Total listing costs divided by total sales =
cost per sale
* 35 States that report all new & used car sales
Case Study: Cars.com ROI Conclusion

cost-per-sale 7.7 X more cost effective than the industry average**
Cost Per Sale
$493*
$68
Cars.com
*Source: NADA 2005
**Dealerships all media channels
Industry Average
Marketer Site Metrics (Lift)
 Marketers often improve their branding among consumers vis-à-vis the
favorable environments provided by the publishers
Metric
Ford Rent-a-Car on
Yahoo! Local
International Bank on
CareerBuilder.com
Accounting Firm on
CareerBuilder.com
Favorable Overall
Impression
7%
10%
-1%
Likely to Use Again
12%
10%
4%
Likely to
Recommend
15%
11%
-1%
Is a Name I Trust
10%
9%
-2%
Lift=Metric (Test)-Metric (Control)
Q (Ford Rent-a-Car, Bank, Accounting Firm, Cars.com): What is your overall impression of_________?
Q (Ford Rent-a-Car): How likely are you to use _____the next time you rent a car?
Q (Bank, Accounting Firm): How likely are you to search the _____job listings the next time you search online for a job?
Q (Cars.com): How likely are you to search on Cars.com the next time you search online for a new or used vehicle?
Q (Ford Rent-a-Car, Bank, Accounting Firm, Cars.com): How likely are you to recommend_________ to a friend, family member or colleague?
Q: Please indicate the extent to which you agree or disagree with the following statements, Attribute rating: I trust this brand.
Base=Respondents aware of marketer
Other Ad Effects (Lift)
 Even ad recall, a high hurdle for an online listing, jumps among test
consumers in a few select cases
 Marketers can impact ad recall if focused on the limited creative element
 Most respondents who visit these sites take some desired action – few are
“just browsing.”
Other Effects
Ford Rent-a-Car
on Yahoo! Local
International
Bank on
CareerBuilder.com
Accounting Firm
on
CareerBuilder.com
Dealerships on
Cars.com
Recall Online
Advertising
-1%
5%
11%
17%
Took Action1
91%
76%
91%
55%
Lift=Metric (Test)-Metric (Control)
Notes 1-Asked of Test Group respondents only;
Q (Ford Rent-a-Car, Bank, Accounting Firm): Where have you seen or heard the following _________advertised or promoted in the past 6 months?
Q (Cars.com): Where have you seen or heard the following _________advertised or promoted in the past 12 months?
Q (Ford Rent-a-Car, Bank, Accounting Firm, Cars.com): What action did you take, if any, after visiting _________?
Base=Respondents aware of /used marketer
Other Ad Effects (Lift)
Most recall Verizon Superpages.com online advertising.
Overall, Verizon Superpages.com experiences an 18 point lift across the board.
Other
Effects
Real Estate
Professional
Services
Automotive
Trade
Services
Hospitality
Retail
Recall
Online
Advertising
14%
19%
16%
9%
19%
23%
Lift=Metric (Test)-Metric (Control)
Q : Where have you seen or heard the following sites advertised or promoted in the past 6 months?
Base=Respondents aware of /used Verizon Superpages.com
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Agenda
Overall Study Objectives
Methodology Review
Summary of Results
 Addressable Market and Opportunity
 Ad Effectiveness Summary
• Conversion
• Return on Investment
• Other Advertising Impacts
 Value of Consumer Target
Key Conclusions and Implications
Key Conclusions

Online Classifieds and Local Advertising is a highly effective form
of advertising
 The case studies show positive ROI

The majority of the value of online advertising is realized via
offline sales
 Tracking offline conversion is critical to fully understand ROI
 Secondary branding impacts are often present

User experience is critical to converting consumers
 Significant attrition occurs in the case
 Marketers must take care not to neglect the online use experience
Key Conclusions (continued)

Marketers must recognize the size and scope of the online local
marketplace
 All forms of local advertising reach a significant portion of consumers
 The channel “flow” may be reversing – online pushing to offline?

The local space attracts highly valuable consumers
 More likely to convert – online AND offline
 Higher income
 More likely to spend time online than average (which means they are
spending less time in offline channels)
Key Implications

Classified and local search listings must offer consumers multiple
ways to complete sale
 40% of Ford’s car rentals were completed offline…by phone, email or
dealer address provided in listing

Classified and local search listings are not just about moving
product today, but about impacting brand image
 Ford and an international bank saw 15% and 11% increases
respectively in “likelihood to refer brand to a friend”
 Ford and an international bank saw 10% and 9% increases in
“trustworthiness of brand”

Classified and local search listings must sell not just list
 Ads must provide product/service benefits and availability
Thank you
IAB Research
comScore Networks
Doron Wesly
Erin Hunter
Director, Industry Research
Senior Vice President
+1 (212) 380-4714
+1(650) 244-5408
[email protected]
[email protected]