Cost Effectively Reaching The In Market Auto Buyer1

Cost-Effectively Reaching the
In-Market Auto Buyer
Alternatives for Tier 1 and Tier 2 marketers to deliver
online advertising performance using networks

An Online Advertising Research White Paper
By ValueClick Media and Goodway 2.0
November 2007


Table of Contents
Executive Summary 3

The Rationale for Auto Sites 4

Introducing the Alternative 5

National Campaign Example (Tier 1) 6

Regional Campaign Examples (Tier 2) 7

Regional Campaign #1 8




Unexpected Results 11

Other Relevant Tactics for Auto Marketers 14

Conclusion 15

About the Authors 16

©2007 ValueClick, Inc., All Rights Reserved 2


Executive Summary
There is a commonly-held belief that third party automotive information sites provide the best channel
to reach in-market auto buyers. While these sites offer attractive inventory proven to drive leads, low
inventory levels and high rates make this one of the most expensive online channels for auto marketers.
This white paper presents the ad network model as one alternative, by using data from five auto cam-
paigns to demonstrate that, when used effectively, the extensive reach and optimization capabilities of
networks can deliver in-market auto buyers on a more cost-effective basis.

Purpose
The purpose of this study, conducted by ValueClick Media and interactive agency Goodway 2.0, was to
evaluate networks as an alternative to advertising on automotive information sites. Moreover, the study
sought to prove the most effective way to utilize networks, namely that running on a large number of
sites and allowing optimization technology to determine those which are most effective will yield higher
performance than presupposing what sites or channels will generate an advertiser’s intended results.

Study Data
Four regional campaigns run on the ValueClick Media network between March and September 2007:
Chevrolet, Dodge, Honda, Lincoln Mercury
One national campaign run on the ValueClick Media network in July and August 2007:
Hyundai

Methodology
The national campaign compared click-through performance of sites that were pre-targeted and consid-
ered likely to reach the intended target audience with a run-of-network media plan that was optimized
based on real-time campaign data and analysis. The regional campaigns studied overall click-through
and conversion rates, as well as performance of individual categories, to draw conclusions about the
cost-effectiveness of using the network model to reach in-market auto buyers. Clients were also sur-
veyed and asked to predict what categories would deliver the best click-through and conversion rates.
This data was compared to actual category performance to illustrate the difficulty in pre-determining
what categories of sites will perform best for any given campaign. Additionally, third party data from
J.D. Power & Associates (2006 Online Cross-Shopping Study) and comScore (Plan Metrix winter 2007
data) was used to better understand auto site visitors and the online behavior of in-market auto buyers.

Key Findings
• The overwhelming majority of auto site visitors are not in-market to buy a car, and those who are
spend less than one percent of their online time on these sites.
• All five campaigns studied revealed that ad networks can be used to cost-effectively achieve
the performance objectives of auto marketers by starting from a broad media plan and optimizing
down to the specific sites that demonstrate proven performance.
• The channels and sites assumed to reach in-market auto buyers are not always the best performing
ones, and have proven to often be detrimental to campaign performance.



The Rationale for Auto Sites
Automotive marketers are generally segmented into Tier 1 (original equipment manufacturers), Tier 2
(dealer associations) and Tier 3 (local dealerships). While Tier 1 auto marketers are concerned with both
brand and direct response objectives, Tier 2 marketers are largely focused on targeting the “in-market
automotive buyer,” which is typically defined as someone who is ready to purchase a vehicle within
four to six weeks. This paper addresses Tier 1 and Tier 2 marketers and their mandate to turn in-market
automotive shoppers into buyers.

It is understandable that auto marketers first think of third party automotive information sites (also
referred to as shopping sites, resource sites, or auto sites), as they offer attractive inventory often posi-
tioned directly in line with impartial reviews of their brands and models. These properties include sites
like Autobytel.com, AutoTrader.com, Cars.com, CarsDirect.com, Edmunds.com, Kelley Blue Book and
NADA.com as well as the auto channels of portals like AOL, MSN and Yahoo. Based on the content of
these sites, it would be fair to assume a close association between these properties and in-market auto
buyers.

However, this assumption has been proven wrong; a 2006 Online Cross-Shopping Study from J.D. Pow-
er & Associates reports that 83 percent of visitors to third party automotive shopping sites do not intend
to purchase a vehicle within the next three months (as indicated by Figure 1 below). More specifically,
only nine percent intend to purchase within a month and only eight percent intend to purchase within
one to three months. Further, 45 percent are not even in the market at all and described themselves as
“just looking.”

Figure 1: Auto Site Visitors’ Intent to Purchase



“While third party automotive As observed by Gene Cameron, executive director, media solutions at J.D. Power
shopping sites have proven & Associates:
their ability to deliver quality
“This very real phenomenon – of in-market buyers making up a relatively small per-
leads, there is sound logic in
centage of auto site visitors – underscores the fact that consumers engage in a wide
exploring the performance
range of activities spanning the entire Internet during the time they are in market to
driving capabilities of other, purchase an automobile. While third party automotive shopping sites have proven
supplemental inventory like their ability to deliver quality leads, there is sound logic in exploring the performance
that which is available on ad driving capabilities of other, supplemental inventory like that which is available on
networks.” ad networks.”

Additionally, in-market auto buyers spend very little time on auto sites compared to the time they spend
on all other online properties. comScore Plan Metrix (winter 2007) data shows that people who have
identified themselves as very likely to buy an auto in the next six months only spend 0.57 percent of
their time online on automotive resource sites such as AutoTrader.com, Kelley Blue Book and the major
portals. Six months go well beyond the in-market window, which is normally considered to be four or
six weeks, giving a clear idea of how little time – relatively speaking – in-market buyers spend on auto
sites. Further, to draw an analogy, dealer associations who buy television advertising do not insist on
solely buying on automotive-related programming; they also spend significant amounts on lifestyle
programs. Correspondingly, high-performing sites on which to advertise to reach in-market auto buyers
often have nothing to do with automotive content.

Still, marketers trying to reach in-market auto buyers pay a premium for this inventory. While it may
seem logical and even “safe” to buy third party auto sites - and they certainly have a prominent role in
the online marketing mix - as advertisers become savvier and more in-tune with the value online met-
rics can provide, decisions that appear to be “safe” should be challenged.

Introducing the Alternative
With such a low percentage of auto site visitors actually in market to buy, and because those who are
in fact in-market spend very little time on those sites relative to their total time online, automotive
marketers should explore other alternatives. To illustrate the efficacy of the ad network model as one
alternative, ValueClick Media studied four regional campaigns in conjunction with interactive agency
Goodway 2.0 for Chevrolet, Dodge, Honda and Lincoln Mercury to determine the difference in category
performance and the importance of not assuming the content with which in-market buyers are associ-
ated. Additionally, a side-by-side comparison was set up for a national Hyundai campaign.



National Campaign Example (Tier 1)

With one portion highly pre-targeted (only running on 92 sites in the ValueClick Media automotive chan-
nel) and another running across the entire ValueClick Media network and then optimized based on per-
formance over a 30-day period, this national Hyundai campaign (placed by its agency of record) was a
true side-by-side comparison. Early in the second week, it became clear that the two portions performed
differently. More specifically, as performance was closely monitored, the portion of the campaign that
was not pre-targeted could be optimized, and the click-through rate of that portion improved notice-
ably throughout the campaign while the performance of the pre-targeted portion actually decreased (as
indicated by Figure 2 below).

Figure 2: Pre-Targeted vs. Optimized Portion – Daily Click-Through Rates



Moreover, the average click-through rate for the optimized portion of the campaign was more than
twice as high as for the pre-targeted portion.

Campaign Portions Impressions CTR

Pre-targeted portion 1,744,991 X

Optimized portion 3,620,054 2.4X

Figure 3: Pre-Targeted vs. Optimized Portion – Average Click-Through Rates

Regional Campaign Examples (Tier 2)

Of the aforementioned auto brands for which regional campaigns ran – Chevrolet, Dodge, Honda and
Lincoln Mercury – three of them tracked both click-throughs and conversions. Each campaign is dis-
cussed in greater detail below, and the bubble charts for them show surprising performance data for
several of the 21 channels (categories) on which the campaigns ran. The fourth campaign tracked only
click-through rates, but it ran twice and therefore includes other interesting findings.



Regional Campaign #1

This campaign ran for a pickup truck across multiple states in the Southeastern United States. Conver-
sion is here defined as “printing out a test-drive certificate.” As Figure 4 below shows, Careers & Edu-
cation generated the best blend of performance, including click-through and conversion rates, which
came as a surprise to all the partners involved in this project. Another surprise was that Dating & Social
Networking was the highest-performing conversion category. Even more surprising to some, though, was
that Autos & Transportation had a conversion rate dramatically lower than the average of the campaign
while the click-through rate was only slightly higher than the average.

But herein lies the point; guessing where in-market auto buyers can be found and sticking to that
without looking at any campaign data will not lead to campaign success. Further, the true benefit of this
data for the agency and marketer is not only in understanding this campaign’s performance, but also in
being able to use the data to more accurately target their online advertising investments going forward
– both within the current campaign and going into the next one. By removing the lowest performing
categories, performance will automatically increase.

Figure 4: Click-Through and Conversion Rates by Category and Associated Volume

The size of the bubbles represents
the volume in terms of impressions
served for each category

The placement of the bubbles
represents the combined
performance in terms of click-
through and conversion rates

The vertical dashed line represents
the average click-through rate for
the campaign

The horizontal dashed line
represents the average conversion
rate for the campaign




This campaign ran for a different pickup truck and instead, in Texas, and conversion is here defined as
“providing personal information to a dealer,” or “watching the one minute and thirty second video in its
entirety.” Figure 5 below shows that Science and Nature performed better than any other category in
terms of click-throughs while Money & Finance outperformed all categories in conversions. This is inter-
esting for two reasons. First, Money & Finance is traditionally thought to be a fit for luxury vehicles, not
a full-size pickup. Second, traditional thinking tells us that consumers within Money & Finance sites
are “time-challenged” and “on the go.” Yet this group spent the full 90 seconds - more than any other
group - watching a video about a full-size pickup truck. Clearly, stereotypes continue to be challenged.

Additionally, the Autos & Transportation category performed below the campaign average both in terms
of click-through and conversion rate. The assumption that a category consisting of automotive enthu-
siasts is the best way to reach consumers who are ready to buy a car within the next four or six weeks
proves to be incorrect. It is interesting to note that these types of auto sites are sometimes even re-
ferred to as “in-market” sites when real campaign data continues to suggest that in-market buyers can
be more successfully found on other sites in other categories.




the campaign





This campaign ran for a popular family sedan in Atlanta, Orlando and Miami, and conversion is here de-
fined as “clicking out to an individual dealer web site.” As Figure 6 below shows, Money & Finance was
again a top performer, again in terms of click-through rate. Actually, Money & Finance is generally a top
performer in any automotive campaign as those who can afford new vehicles in particular are typically
upper income individuals. Also, Travel and News & Reference are two interest categories known to skew
toward upper income and education, which makes sense using this rationale. However, the success of
Health & Fitness, Music & Radio, and Computers & Software can simply not be explained by the same
idea. Continuing the trend from the other campaigns analyzed above, the fact is that for any product in
any market in any given season there will always be surprises in online results for which the campaign
could not have been pre-targeted.

This is where the exact value of running across a large network of sites in multiple channels lies; while
the early phase of a campaign may be spent learning the behaviors of your product, audience, and on-
line behavior, subsequent spending has a strong likelihood of performing significantly higher than cam-
paigns targeted to the assumed “right” channels. And to address the Autos & Transportation category
again, although its conversion rate was slightly higher than the campaign average, the click-through rate
was lower.




the campaign





The fourth regional campaign only tracked click-through rates; however, this campaign ran twice – first
over a 26-day period in March and then over a 40-day period from July to September 2007. The click-
through rate increased by 85 percent from the first time it ran to the second, further demonstrating the
value of optimizing campaign placements based on actual behavior of the audience. It should be noted
that the creative was different for the two periods, but the click-through rate of the portion of the cam-
paign that ran on the ValueClick Media network outperformed the average click-through rate of other
network placements by 23 percent during the first period and by 29 percent during the second period.
This improvement can be attributed to a focus on removing the categories that did not perform in
March and narrowing in on the ones that did. By learning behavior this way, marketers can reach their
audience in a significantly more cost-effective manner than by repeatedly running ads on the same sites.

Unexpected Results

Client and/or agency representatives for the regional campaigns were asked to force-rank the 21 Value-
Click Media channels (categories) on which their campaigns ran based on their expected performance.
Figure 7 shows the predicted vs. actual ranking of these categories in terms of click-through rate for
one of the campaigns while Figure 8 does the same for conversion rate for the same campaign. As indi-
cated by the dark blue bars, Autos & Transportation was expected to be the top performer while Gaming
was expected to be the worst. However, as the light blue bars indicate, the actual ranking was very
different. The purpose of illustrating this dramatic difference between what the client and/or agency
representatives predicted and the actual results is simply to bring attention to the importance of not
presupposing where a certain audience is likely to be found. The performance of some of the categories
came as a surprise even to the authors of this paper.



Figure 7: Category Click-Through Rates – Predicted vs. Actual Ranking



Figure 8: Category Conversion Rates – Predicted vs. Actual Ranking



Other Relevant Tactics for Auto Marketers
The material covered in this paper addresses specific attributes of the ad network model, namely opti-
mization technology. For automotive marketers who may not have worked extensively with ad networks
or online agencies, it is worth noting additional ways in which to leverage their capabilities.

Transparency – The largest and most reputable networks offer a high degree of transparency and flex-
ibility that should genuinely overcome concerns over ads running next to objectionable content. Some
networks will disclose every site on a plan or offer a list outside of which your ads are guaranteed never
to appear.

Road Blocks – While technically networks are unable to guarantee 100 percent share of voice, they do
have the ability to quickly deliver a high volume of impressions to specific audience segments, creating
high impact that is ideal for vehicle product launches.

Behavioral Targeting – Ad networks are continually improving their ability to identify specific audience
segments based on user behavior, the most prominent of which is re-targeting consumers who recently
visited a manufacturer’s site when they appear on sites within the network.

Content Integration and Sponsorship – Some networks have representation agreements with publishers
that allow them to go beyond serving banner ads to offer sponsorship, promotions and content integra-
tion with higher response rates and unique branding opportunities.

Video – In addition to their longtime ability to run Flash-based in-banner video ads, most major net-
works now offer in-stream video capabilities, creating new ways to extend television assets and online
video strategies.

Lead Generation – Beyond lead generation as a marketing objective, the discipline of generating quali-
fied leads as consumers complete registration forms online is offered by some networks and allows
automotive marketers to quickly build databases of consumers who have requested product information.



Conclusion
Although advertising on automotive information sites is a logical choice for most marketers aiming to
reach in-market auto buyers, this inventory is limited and therefore sold at a premium. Third party re-
search also shows that an overwhelming majority of auto site visitors are not in-market to buy a vehicle
and the ones who are spend less than one percent of their online time on those sites. Further, data from
five different campaigns suggests that a distributed buy performs significantly better in the long run
than contextual buys in terms of both response and conversion.

Ad networks rise out of these findings as a natural option for marketers whose mandate is to convert
automotive shoppers to buyers. With their reach and optimization technology, large ad networks can run
campaigns with broad media plans and optimize down to the specific sites that perform best for each
campaign. Although this concept is far from new, analyzing the category performance from the cam-
paigns in this study makes it clear that pre-supposing where an audience exists and pre-targeting based
on those assumptions will not deliver desired performance.

While targeting and optimization must work interdependently and in varying degrees of emphasis
depending upon the campaign objective, pre-targeting too narrowly prevents campaigns from having the
ability to be optimized. And in general, the stereotype of a web site and its correlation with its users is
rarely a good match. The real-time data provided by ad networks tells the real story of the categories
that drive verifiable success, which is information both Tier 1 and Tier 2 marketers can use to cost-ef-
fectively reach the in-market auto buyer.



About the Authors
Contributing Authors

Jay Friedman, president, Goodway 2.0
Tony Winders, vice president, marketing, ValueClick Media
Joergen Aaboe, marketing specialist, ValueClick Media

About ValueClick Media

ValueClick Media is one of the world’s largest and most effective performance advertising networks. Offering
innovative solutions to advertisers through its advanced display and lead generation platforms, ValueClick Me-
dia leverages its global reach, extensive targeting capabilities and market-leading optimization technology to
accomplish any online marketing objective. For publishers, ValueClick Media offers comprehensive solutions
to earn the highest revenue for their available advertising inventory. Underlying all of ValueClick Media’s solu-
tions is an experienced and passionate team dedicated to helping advertisers and publishers achieve online
advertising results.

30699 Russell Ranch Road, Suite 250, Westlake Village, CA 91362, 818.575.4500, valueclickmedia.com

About Goodway 2.0

Goodway 2.0 is a division of the Goodway Group, a marketing and promotions company originally founded
in 1929 by Milton Wolk. Milton’s Grandson, David Wolk, is now president and owner and formed Goodway
2.0 with Jay Friedman in 2006. Goodway 2.0 has quickly become a leader in the digital space, working with
clients from major domestic and import automotive manufacturers to local and national political candidates.
Each product in the Goodway 2.0 proprietary solutions suite – Niche Web™, DARCO™, Vernon™, Behavioral
Learning™, RateJammer™, and Periscope™ - each offers a unique ability to target consumers online with
increased efficiency and ROI. These products, coupled with extensive retail experience, allows Goodway 2.0
to bring custom solutions to clients who are looking for new ways to reach a digital audience.

Penthouse Suite, The Pavilion, 261 Old York Road, Jenkintown, PA 19046, 215.887.5700, goodwaygroup.com


Cost Effectively Reaching The In Market Auto Buyer1

More Related Content

What's hot (9)

Viewers also liked (7)

Similar to Cost Effectively Reaching The In Market Auto Buyer1 (20)

More from Automotive Social Media Marketing Reputation Management (20)

Recently uploaded (20)

Cost Effectively Reaching The In Market Auto Buyer1