read and analysis

i tell i pick

Air France Internet Marketing:
Optimizing Google, Yahoo!, MSN, and Kayak Sponsored Search

Rob Griffin, senior vice president and U.S. director of search for Media Contacts, a media
solutions provider, examined a report containing this month’s numbers for the profitable paid-
search campaign his agency had managed for one of its top clients, Air France. Executives at the
French company were pursuing an international growth strategy and were looking to increase
their share in the hyper-competitive U.S. air travel market. Although they were pleased thus far
with the results that Griffin and his team had delivered during the engagement, the pressure was
on Media Contacts to continuously optimize performance and return on advertising (ROA)
dollars spent for search engine marketing (SEM) campaigns.

Sharon Bernstein, director of insights for Media Contacts, had recently briefed Griffin on the
situation:

Air France’s revenue from online ticket sales continues to grow. Our reports demonstrate
Media Contacts’ success at driving a growing volume of visitors to Air France–affiliated
Web sites. As we continue to make decisions regarding allocation of funds toward search
marketing, my team is eager to focus our analytic marketing expertise on increasing Air
France’s net revenue gained through online advertising as well as ROA. As you know,
the cost-per-click of search engine keywords is continuing to increase and there are more
new players entering the market. I would like to see our campaigns be even more efficient
at driving visitors to Web sites and converting them to customers while keeping click
costs minimized.

As Griffin digested Bernstein’s comments, he thought about how SEM had become an
advertising phenomenon, with North American advertisers spending $9.4 billion in the SEM
channel in 2006—62 percent more than they had spent in 2005.1 But Griffin did not take this
bullish growth for granted; although he was pleased with his company’s performance to date,
Griffin wanted to make sure that the team kept its edge and could deliver the results Air France
expected. Griffin agreed with Bernstein that a more effective mix of tactics could improve the
efficiency of Media Contacts’ campaigns and continue to deliver positive net revenue and ROA
for Air France. Griffin knew the data Media Contacts had been collecting on its search campaigns

for Air France could be analyzed to optimize future campaign performance; now he had to decide
how to move forward.

Air France

In 1933 five French airlines merged to form Air France, and the new company set up its hub
at the main airport of the Ile de France Region in Paris. Seventy-five years later, that hub, Paris–
Charles de Gaulle, had grown to be one of the most efficient in Europe, providing the largest
number of connections within the shortest time span.

After the end of World War II, Air France launched its first successful flight to the United
States; this inaugural flight between Paris and New York City was held on July 1, 1946, with a
flight time of 19 hours and 50 minutes. By 2007 Air France had expanded its U.S. cities served to
thirteen. In total that year, Air France served 185 destinations in 83 countries and its total fleet
size stood at 383 aircraft.

On May 5, 2004, Air France and Netherlands-based KLM, two of Europe’s oldest airlines,
joined forces and created the SkyTeam global alliance. This alliance was in response to the need
for a global travel offering (Exhibit 1). In 2007 SkyTeam employed a work force of 150,000.
Combined, the two airlines carried the most scheduled international passengers, flying to 225
destinations in 109 countries (Exhibit 2). The new Air France-KLM group developed its strategy
based on a concept of “one group, two airlines.”

As Air France grew, it developed an approach to aircraft fleet management that was based on
a two-prong strategy: rationalization, through acquisition of modern aircraft with similar
technical characteristics (i.e., “family effects”), and flexibility, to adjust aircraft delivery dates or
change models within a given aircraft family. Air France needed to be able to adapt quickly to the
changing demands of its environment. If demand was low, the company had to reduce capacity
by eliminating unnecessary aircraft. For this reason, a substantial part of Air France’s fleet was on
short- or medium-term lease, a practice known as a “progressive operating lease.” Thanks to this
strategy, Air France registered positive results from 1999 through 2007, despite years of crisis in
the airline industry.

By the end of the 2006–2007 fiscal year, Air France-KLM continued to see growth. During
this time, the global economy grew by 4.9 percent, while airline traffic increased 6.6 percent,
according to the Association of European Airlines. At Air France specifically, passenger activity
increased 5 percent, with 73.5 million passengers carried, resulting in Air France posting a 5.1
percent rise in unit revenue per available seat kilometer.

The Airline Industry and Its Competitive Landscape

Profitability is finally on the horizon. Airlines can be proud of impressive results in
safety, efficiency, and the environment. But profits still don’t cover the cost of capital. Air
transport must continue its agenda for change.

—Giovanni Bisignani, director and CEO of the
International Air Transport Association (IATA)

The airline industry had historically been plagued with low returns, bankruptcies, and ever-
fluctuating demand. But in 2006 airline profitability finally improved. After six challenging years
and $42 billion in losses, airlines came within $500 million—0.1 percent of revenue—of what
industry analysts considered breaking even. Furthermore, operating profit for the industry reached
$13 billion in 2006, more than double the amount generated the previous year. At just over 3
percent of revenues, however, this was still far from ideal industry profit levels (Exhibit 3).

Nearly all of the $42 billion in losses accumulated in the six years prior to 2006 had been
generated in the U.S. market. In 2006 the U.S. airline industry returned to profitability, excluding
the $10 billion cost of bankruptcy restructuring. Outside the United States, airlines had suffered
net losses in 2001 following the events of September 11, but in subsequent years, most airlines
had seen the return of modest profits.

International travel had begun to emerge as the fastest growing market; in 2006 the number of
international passenger kilometers grew by 6 percent, while growth in domestic passenger
kilometers was less then 4 percent. Although two-thirds of air passenger journeys were domestic,
when measured in revenue passenger kilometers (RPK), international travel accounted for 60
percent of airline revenues.

In addition to the general increase in demand for air travel, from 2000 to 2006 the industry
saw a 50 percent rise in the number of passengers traveling on economy tickets while travel on
premium tickets remained stable. In 2006 international premium traffic grew 4.3 percent, much
less than the 7.4 percent growth in economy traffic.

Taking advantage of this growing demand for economy air travel was Ryanair, Europe’s
original low-fare airline. In 2007 the company led the market for international air travel. The
airline offered services across twenty-six European countries but had not expanded its offerings
to the United States. French-operated L’Avion, on the other hand, catered to the traditionally less
price-sensitive business traveler. Indeed, in 2006 L’Avion emerged onto the international airline
scene as a provider of business-class-only service between Paris and Newark. This route was the
airline’s sole offering that year.

Other airlines operating in the Air France market space included U.S. carriers American
Airlines and United, as well as European carriers Lufthansa and British Airways. At the time,
American Airlines carried more passengers than any other airline.

Transformation of Travel Industry by the Internet

World Wide Web and the Emergence of E-Commerce

The travel industry was one of the earliest to adopt e-commerce into their sales strategies.
Driven primarily by the development of the World Wide Web, the Internet had evolved from its
origins as an institutional and educational network into a consumer-driven network. The user-
friendly interface of Web-browsing software quickly led to development of consumer-geared
Web sites. Consumers could access information more readily and at their own convenience. This
newly found accessibility held huge potential for increasing businesses’ sales by vastly increasing
their reach. As consumers grew more Web-savvy and became more trusting in the security of
online services, e-commerce emerged as a very fast growing sales medium.

Having already built systems for external use by travel agents to search and book flights for
customers, airlines were uniquely well suited to establish direct-to-consumer sales portals via the
Internet. Further, with the adoption of e-tickets, airlines were better protected from many of the
logistical problems faced by other industries as they evolved toward online sales.

The Landscape of the Travel Industry on the Internet

The travel industry comprised several different types of online service providers. Direct Web
sites were owned and hosted by the individual airlines, and as popularity of the Web and e-
commerce grew, travel providers responded quickly by establishing their own Web sites.
Airlines’ direct Web sites provided a number of services beyond the ability to purchase flights,
which made them attractive destinations for travelers performing their own Web searches. These
Web sites housed full flight schedules for the airline, real-time information on flight
arrival/departure times, fleet information, and customer loyalty programs. As a result, the number
of user-oriented travel tools on airlines’ direct Web sites increased dramatically in just a few short
years (Exhibit 4).

Consumers could also choose to purchase air tickets from aggregator Web sites such as
Expedia.com, Orbitz.com, and Priceline.com. The aggregators existed in a pseudo-competitive
position relative to the airlines’ own direct Web sites but also offered services beyond flight
purchase, including hotel and vacation packages. When consumers would book via the
aggregators, the airlines still earned revenue, but they paid a fee to the aggregator for facilitating
the purchase. The airlines’ direct Web sites had difficulty competing with the convenience of the
aggregators, which could perform lowest-price searches and allowed the consumer to build full
travel packages with the convenience of one purchase from one Web site.

Further increasing the number of travel-booking alternatives online was a third type of
service provider, metasearchers. The metasearch sites (e.g., Kayak.com, Sidestep.com) were also
aggregators of information but did not offer transaction services. Instead, consumers could use
these sites to search for their travel services, view consolidated results from multiple direct
service providers, make their choice, and then be linked through to the airline’s direct Web site to
complete their purchase. Travel providers paid these sites for advertising (in a similar fashion to
the major search engines) in order to be included in the search results.

Consumer Adoption of E-Commerce for Travel

According to the Travel Industry Association (TIA), the number of consumers going online
to research and book travel increased 263 percent between 1996 and 2005.2 All estimations of the
online travel market predicted substantial growth rates. PhoCusWright, a travel industry
marketing research authority, estimated that more than one half of all U.S. travel would be
booked online by consumers in 2006, compared to 30 percent in 2005, 20 percent in 2003, and 15
percent in 2002.3

The intangibility of travel as a consumer product made it uniquely well suited to online
purchasing when compared to other segments of the retail industry. With other products,
consumer decision making could be difficult without handling the product prior to purchase. This
was not the case with travel purchases. Additionally, increased access to the Internet by
consumers, brought on by greater availability of high-speed connections, helped bolster this trend
considerably, along with increased price competition among suppliers.4 While overall Internet
penetration approached saturation, the volume of high-speed connections was continuing to rise
(Exhibit 5). In 2005 an estimated 32.2 million households in the United States had high-speed
Internet connections.5

Search Engine Marketing

With billions of Web pages to sort through, search engines provided the primary means for
navigating and organizing the Internet. SEM became a well-known method of marketing in which
businesses promoted their products and services through targeted placements on Internet search
engine results pages (SERPs). Search engines constructed SERPs using proprietary algorithms,
which determined the most relevant sites for a given search. This provided a highly targeted space
to connect consumers to exactly what they were looking for when they were looking for it. In
2006 SEM had become the most successful form of online advertising because of this targeted
relevancy. SEM involved both Web site search engine optimization and pay-per-click sponsored
search campaigns.

Search Engine Optimization (SEO)

SEO analyzed the structure and content of a business’s Web site to maximize its readability
and relevance to search engine “robots” indexing the content of the Internet. SEO strategies
looked at technical aspects of sites such as URL address structure, Web server settings,
information architecture, site usability, and text content. Image- and multimedia-heavy sites
required adding alternate text tagging to be readable by search engines. The goal of SEO was to
organically improve a site’s relevancy ranking, which meant it would naturally appear higher on a
SERP.

Pay-Per-Click, or Sponsored Search

In pay-per-click, or sponsored search, advertising campaigns, businesses bid on keywords for
sponsored link listings consisting of a title, a short description and a display URL. Sponsored
links usually appeared at the top of a SERP for a relevant search. Relevancy was determined by
the match of keywords bid on by the advertiser, the amount bid for those keywords in comparison
with competitor bids, the number of times an ad had been clicked on for a keyword, and
proprietary “black box” algorithmic criteria from the search engine provider. Generally, an
advertiser paid for an ad only when it was clicked. This cost-per-click could reduce with volume
as the search engine provider determined the ad was more relevant to those keywords.

Sponsored search could be managed in terms of campaigns, keywords, and groups of
keywords. Although each search engine had unique terms for its programs, most had become
similar, likely due to the success of Google’s program. Advertisers placed maximum bids on
keywords and laid out daily budgets for campaigns. These budgets, along with relevancy,
determined how much share-of-voice (SOV) the advertiser achieved. The larger search engines
allowed for broad, exact, and phrase keyword campaigns to allow advertisers to cast either wide
or narrow nets of impressions. Choice of keyword type depended on the uniqueness of a keyword
to describe the underlying product or service. Geotargeting, a relatively new search engine
feature, allowed a campaign to target a specific geographic region and language.

Emerging trends in sponsored search tended to revolve around more finely targeted
relevancy. Search engine verticals had appeared in several industries, such as Kayak for travel
services, to allow for more specific searches within the given industry. Local search had also
achieved buzz as a means to target consumers with locally relevant search results; geotargeting
was a form of local search.

Pay-per-click campaigns were considered easily measurable, due to the ability to track
precise actions on the Internet. Using analytic software, an advertiser knew the number of times
an ad was seen (known as impressions), which keywords were used, the number of clicks per
impression or click-through rate on the ad, how many clicks were converted into transactions
(known as transaction conversion percentage), and how much revenue those resulting transactions
were worth.

To optimize a sponsored campaign, the campaign had to improve on one or more of the
following: cost-per-click reduction, increase in bookings, net revenue, revenue per transaction,
return per transaction, overall performance by engine, or other performance metrics that could
improve the net revenue of a particular campaign.

Sponsored search provided more immediate traceable results than SEO and generally was
more campaign driven. The instant tracking of results and on-the-fly adjustments allowed pay-
per-click campaigns to be highly responsive.

SEO and pay-per-click worked together in a SEM strategy: performance of pay-per-click
keyword campaigns could be used to tweak content on the home Web site, and the relevant
portfolio of pay-per-click keywords could be inferred by analyzing which keyword referrals were
most effective at leading a consumer to the Web site through a natural listing.

Media Contacts and Its Partners

In 2007 Air France operated in the fiercely competitive business of international and
domestic airline travel services. Catering to both business and leisure travelers, Air France turned
to Internet marketing campaigns and search engine optimization to reach large customer segments
in multiple countries including the United States. To be successful, Air France had to understand
how to maximize the net revenue and the ROA of its Internet marketing campaigns by evaluating
alternative strategies. The airlines hired Media Contacts to help it achieve this goal.

Media Contacts

Media Contacts was the global interactive media network of Havas Media, a group ranked
sixth among communications consulting companies throughout the world. Beginning in 1997,
Media Contacts had grown to operate twenty-seven offices in twenty-three countries in Europe,
North America, South America, Asia, and Oceania. The company provided data-driven media
solutions across all interactive channels, from direct response to relationship-based media. In
addition to Air France, Media Contacts’ extensive list of customers included Royal Caribbean,
Fidelity, and Goodyear. John Arnott, Web and communications manager of ING Direct (UK),
commented on Media Contacts’ service:6

One of the most refreshing aspects of working with Media Contacts is that they are
genuinely passionate about our brand and delivering results for us. They constantly look
to improve our online performance through innovative thinking and testing new
approaches. The level of analysis and reporting we receive from them is crucial to the
continuing success of our business.

Media Contacts worked to form strategic partnerships with a number of research and
technology providers in order to gain access to granular data sets, which could be integrated with
the company’s campaign data. One of the partnerships Media Contacts established was with
DoubleClick, a company that developed and provided Internet ad services. In 2007 Media
Contacts was one of three agencies in the world that transferred all impressions, clicks, and
activity data from DoubleClick services to its data warehouse on a nightly basis.

DoubleClick

As an Internet ad services company, DoubleClick offered technology products and services
marketed primarily to advertising agencies and media companies in an effort to allow clients to
traffic, target, deliver, and report on their interactive advertising campaigns. The company’s main
product line, known as DART enterprise, was intended to increase the purchasing efficiency of
advertisers and to minimize unsold inventory for publishers.

Included in the DART enterprise system was a Web-based search system, which was
integrated with leading engines such as Google, Yahoo, and MSN. With the increased Internet
use for client purchases, buyers of online advertising struggled to find the best allocation of ad
dollars. DoubleClick’s Advertising Exchange service was designed to help customers maximize
ROA through dynamic pricing and intelligent bidding. The tool allowed search specialists to set
maximum bids based on past performance with specially designed bid rules. In addition,
specialists were able to define budgets and time frames, as well as select cost-per-thousand, cost-
per-click, and cost-per-action pricing models. Customers using the DART enterprise system
spoke highly of its benefits:

DART Enterprise has helped us to understand where users are going on our site and
their buying behavior patterns so we can continually adjust offers, placements, and
creatives to improve buy-through and maximize sales.

—Otto Linton, ad trafficking manager, VEGAS.com

In 1999, at a cost of $1.7 billion, DoubleClick merged with data-collection agency Abacus
Direct, which worked with offline catalog companies. Roles changed for DoubleClick in 2007,
however, when Google announced it had reached an agreement to acquire DoubleClick for $3.1
billion in cash, a decision that quickly stirred anti-competitive concerns regarding Google’s
power and influence in the growing SEM industry.

Google

Google began in January 1996 as a research project by two PhD students at Stanford
University. Based in Mountain View, California, the company was incorporated in 1998 and
received its first round of venture capital funding in June 1999. Google’s advanced Internet
search engine technology positioned the company for rapid growth, and following an initial
public offering in August 2004 that raised $1.67 billion, the company pursued an aggressive
growth strategy through new product developments, acquisitions, and partnerships. By June 2006
Google employed 7,942 full-time workers and was one of the top five most popular sites on the
Internet.

Google generated revenue primarily through two programs: highly targeted advertising and
online search services. Google’s advertising program used auction pricing based on the value
(cost-per-click) an advertiser assigned to particular keywords. The position of a particular
advertisement on Google’s search results page was determined by a combination of the cost-per-
click and the click-through rate so that the most relevant ads were displayed more prominently.
While these tactics allowed Google to display the more relevant ad copies higher on a given page,
Google’s page-ranking technology also utilized certain “black box” algorithms and calculations
to which search engine marketers did not have access.

Powering Google’s search engine technology and advertising services was an advanced
technology architecture that linked large numbers of inexpensive PCs together in a highly
efficient grid network. Employing highly complex mathematical algorithms, Google software
conducted a series of simultaneous calculations requiring only a fraction of a second. Unlike its
early competitors who relied heavily on how often a word appeared on a Web page when
displaying search results, Google used its patented PageRank algorithm to determine which Web
pages were most important. According to Google’s Web site, “PageRank reflects Google’s view
of the importance of Web pages by considering more than 500 million variables and 2 billion
terms. Pages that Google believes are important pages receive a higher PageRank and are more
likely to appear at the top of the search results.”7

Not surprisingly, Google’s growth did not go unnoticed by competitors such as Yahoo and
Microsoft. As consumers and businesses increasingly shifted a larger percentage of purchases to
Internet channels, SEM and digital advertising campaigns represented a significant growth
industry. Google launched the opening volley in this race for market share by announcing in 2007
that it was acquiring DoubleClick. Tim Armstrong, Google’s president of advertising and
commerce for North America, explained the rationale for the acquisition:

This transaction will strengthen our advertising network by expanding our access to
publisher inventory and enabling us to serve the needs of a broader set of advertisers and
ad agencies.

Microsoft MSN

The Microsoft Network (MSN) was released in August 1995 and established Microsoft as a
major player in the online world. Originally focused on Web e-mail and positioning itself as an
Internet service provider, Microsoft realized by 1999 that the largely underused MSN.com
domain name would require a certain level of rebranding to better challenge Yahoo for online
advertising dollars. As a result, that same year MSN.com was relaunched as a Web portal and a
branded family of sites produced by Microsoft’s Interactive Media Group (IMG). Perhaps the
most significant aspect of the relaunch was Microsoft’s decision to offer the majority of content
for free, compared with its previous policy of paid subscription services.

Early in its inception, MSN realized an opportunity to partner with other service and content
providers in order to drive traffic to the MSN.com home page. MSN received revenue from three
key sources: Microsoft adCenter, MSN Shopping, and the subscription-based Internet service to
Web users. Microsoft adCenter (formerly MSN adCenter) was developed internally by Microsoft
to deliver ads and to end Microsoft’s historical reliance on third-party providers. As late as 2006
Microsoft had relied heavily on Yahoo for search-related advertising; now it began creating its
own search product within adCenter.

Despite a comparatively late entry into the SEM world, MSN.com enjoyed at least one
advantage over pure-play search engines: rich customer data gathered through its Internet hosting
service. At one point, MSN.com claimed approximately nine million subscribers, second only to
AOL.com in terms of registered users. By combining the technology capabilities of Microsoft
adCenter with customer transaction information, Microsoft moved to differentiate itself from its
competitors.

Like Google adWords, Microsoft adCenter used pay-per-click technology when serving
advertisements as well as advertisement click-through rate in determining the frequency at which
an ad would be displayed. Microsoft differed from Google, however, by allowing its advertisers
to target ads to a particular customer demographic. In other words, by restricting ads to a given
set of demographics, Microsoft could increase the bid price whenever users of a particular
demographic would see an ad.

Despite adCenter’s promising potential, Microsoft was keenly aware of the threat posed by
Google. When Google announced its intention to acquire DoubleClick, Microsoft responded with
a $6 billion tender offer for digital marketing firm aQuantive—an offer that represented an 85
percent premium over aQuantive’s stock price in the days preceding the announcement. When
Kevin Johnson, president of the platforms and services division at Microsoft, announced the
acquisition, he erased any doubt of Microsoft’s perspective of Google:

The Microsoft aQuantive transaction will promote competition, and the Google
DoubleClick transaction will reduce competition . . . Consider on the one hand that
aQuantive today is in three lines of business . . . Microsoft today is in none of those
businesses. That’s why this acquisition will increase competition.8

Yahoo

In early 1994, Stanford graduate students Jerry Yang and David Filo created a Web site that
would later become known as Yahoo. Originally designed as a directory of other Web sites,
Yahoo in time shifted course, evolving into a searchable index of Web pages. As it grew
organically in size and breadth of the products it offered, Yahoo expanded into new markets by
acquiring a broad mix of companies. With each acquisition, Yahoo modified the terms of service
for customers of the acquired company, claiming intellectual property rights for the digital
content residing on Yahoo’s computer servers. By leveraging this content, Yahoo pursued a
strategy of expanding beyond pure-play search and becoming a content-driven Web portal for the
global Internet user community.

As the predecessor of Yahoo Search Marketing, commercial Web search services company
Overture was one of the early pioneers in monetizing Internet search traffic. Its May 1999 patent
application, entitled “System and method for influencing a position on a search result list
generated by a computer network search engine,” was used in litigations against several
companies, including Google. Interestingly, it was precisely during this litigation, in 2003, that
Yahoo acquired Overture. Perhaps to hasten the buyout of Overture, the Google vs. Overture
lawsuit was settled the next year, with Google agreeing to issue 2.7 million shares of common
stock to Yahoo in exchange for a perpetual license of the underlying patent’s technology.

Yahoo pursued competitive differentiation in SEM by combining ads with “content
matching.” Content matching represented Yahoo’s ability to marry search and non-search content
into a single unified user experience. Yahoo’s extensive Web portal network and digital content
repositories provided the company with a unique ability to display more than just search results;
context-specific content could be linked to search keywords, a function Yahoo hoped would
present a greater value proposition than pure-play Internet search providers. It should be noted,
however, that Yahoo received 46 percent less revenue per SERP than Google.

While Google and Microsoft had enjoyed measurable success in their respective search
marketing businesses, Yahoo’s performance had been comparably less positive. Yahoo responded
to lackluster results in 2006 by replacing its CEO of six years, Terry Semel, with company co-
founder Jerry Yang. As a first-time CEO, Yang acknowledged the challenges ahead during his
first earnings conference call:

I intend to spend the next 100 days or so focused on mapping out a strategic plan for
long-term success, working with our teams to put the right organization and the right
people in place, and making any necessary changes.9

Kayak

Founded in 2004 by Steve Hafner and Paul English, Kayak.com was a relative newcomer to
the SEM industry. The company was considered a travel aggregator of sorts, meaning it scoured
the Internet for the best possible business and leisure travel services. Kayak’s founders were
experienced entrepreneurs with a history of successful startup businesses under their belts.

Focusing on a specific niche in the SEM industry, Kayak aimed to differentiate itself in two
ways: as an alternative business model from those of established travel service aggregators such
as Orbitz, Expedia, and Travelocity; and as a unique technology architecture. Long-time industry
players such as Orbitz and Expedia purchased and resold airline seats and hotel rooms, in effect
committing to expensive inventory holding costs in a notoriously volatile and price-sensitive
market. Kayak, however, was exclusively a travel search service; rather than sell anything, it
merely connected consumers with the airlines and hoteliers.

The company also intended to leverage highly advanced software application architecture that
was distinct from existing competitor offerings. Instead of passing Web site data through
expensive central servers, Kayak’s rich Internet application (RIA) architecture transferred some
of the processing load to the consumer’s computer, ultimately allowing much greater (and
cheaper) processing capacity. As a result, Kayak said it “provided a more exhaustive search than
any other travel site, including the inventory of 551 airlines and 91,500 hotels. Its goal is to find
and search every hotel on earth, which it estimates at 300,000 properties.”

In addition to the company’s service operations and technology architecture, Kayak.com
enjoyed another distinction relative to the larger search engine providers such as Microsoft,
Yahoo, and Google: Kayak’s Web site search and clickstream data were not integrated with
DoubleClick. As a result, advertising agencies and media companies interested in analyzing
Kayak.com site activity had to rely on alternative data sources and data integration techniques for
an effective comparison. Because of Kayak’s comparably smaller market share position in the
industry, extensive analysis of Kayak data was rare.

Kayak’s historical business results were less than impressive, but the forecast was beginning
to look more positive. Although monthly losses were running at $500,000, sales volume was
increasing at 15 percent per month, reaching $388,000 in September 2005. In addition, the
company boasted a click-through rate of 8 percent for Kayak.com, significantly higher than the
0.8 percent industry average for online travel sites.

Details of the Challenge

The Media Contacts team had only a short time to make sense of all of its research and
determine a way to optimize future campaigns. Although it appeared that using branded keywords
might bring in more revenue, it was also apparent that unbranded keywords produced a larger
percentage of single-click conversions. Which approach would have a higher propensity to
increase ticket sales and to improve ROA? Furthermore, Media Contacts had to determine
whether broad or focused keywords were more profitable for Air France. Although broad
keywords occupied the majority of searches, were they as profitable as focused keywords? How
could Media Contacts use assist keywords to lead consumers to a desired behavior? Note that
typical Internet tracking worked by measuring the last keyword clicked, and this keyword was
credited for the sale/conversion. The team’s research therefore extended well beyond the
examination of numerical data; it had to think about consumer behavior, considering how many
searches it might take for someone to convert.

Each search engine that Air France was working with in July 2007—Google, Yahoo, MSN,
and Kayak—appeared to specialize in different services and consequently might appeal to
different audiences. Media Contacts needed to decide which search engine(s) delivered the most
value to Air France per dollar spent. They also needed to be mindful of whether any search

engines were partnered with travel aggregators, as this could create a potential conflict of interest
for Air France.

Media Contacts could use DoubleClick data to analyze past performance and make
recommendations for strategy optimization. The data included line item information for keywords
and keyword groups for each campaign and search engine (Exhibit 6). To make sense of this
information, it would need to be grouped and analyzed using Microsoft Excel’s pivot table
feature (Exhibit 7). Using pivot tables to show different summaries of data, the Media Contacts
team could analyze campaign key performance indicators (KPI) by search engine, bid strategy,
keyword group, keywords, and so on.

Next Steps

As Griffin and his team reviewed the latest marketing campaign figures for Air France, he
paused to reflect on potential future strategies Media Contacts should consider for increasing Air
France’s market share and profitability in the United States. While a quick review of the data
indicated that SEM was a profitable venture for Air France, Griffin was eager to understand what
a more thorough data analysis would uncover. For example, he was curious whether different
search engines were attracting different consumer segments and yielding unique purchase
patterns and consumer behavior from SEM campaigns. Glancing at his calendar, Griffin noticed
the upcoming biannual financial planning session with the Air France marketing team to review
campaign funding proposals.

• Should Media Contacts recommend a uniform strategy for Air France across search
engine publishers? Or would it be more effective to tailor each publisher strategy to
maximize return on investment?

• How can campaigns be improved to increase overall value gained from investment with a
search engine publisher? Should keywords be added or dropped from the campaign?
Should campaign tactics or copy be adjusted to improve campaign performance?

• What are the most important KPIs, and what impact will campaign changes have on these
KPIs?

• How should future SEM campaigns be structured? In the past, Media Contacts had
concentrated on Google, Microsoft, and Yahoo; was there now an opportunity to
optimize search advertising with metasearch companies such as Kayak?

7%
16%

7%

15%
14%

28%

13% Caribbean & Indian Ocean
North America
South America
Asia
Africa & Middle East
Europe
France

Exhibit 1: Scheduled Passenger Revenue by Destination

Source: World Air Transport Statistics (WATS). WATS provides statistics from more than 350 airlines, including low-cost carriers. This
represents a complete statistical picture of the airline industry in 2006.

A
IR

F
R

A
N

C
E

I
N

T
E

R
N

E
T

M
A

R
K

E
T

IN
G

E
xh

ib
it

2
S

ch
ed

ul
ed

In
te

rn
at

io
na

l P
as

se
ng

er
s

C
ar

ri
ed

in
2

00
6

(W
or

ld
)

S

ch
ed

ul
ed

P
as

se
ng

er
s

C
ar

ri
ed

in
2

00
6

(W
or

ld
)

R
an

k
A

irl
in

e
N

o.
o

f P
as

se
ng

er
s

R

an
k

A
irl

in
e

N
o.

o
f P

as
se

ng
er

s
1

R
ya

na
ir

40
,5

32
,0

00

1

A
m

er
ic

an
A

irl
in

es

99
,8

35
,0

00

2
Lu

fth
an

sa

38
,2

36
,0

00

2

S
ou

th
w

es
t A

irl
in

es

96
,2

77
,0

00

3
A

ir
F

ra
nc

e
30

,4
17

,0
00

3
D

el
ta

A
irl

in
es

73

,5
84

,0
00

4

B
rit

is
h

A
irw

ay
s

29
,4

98
,0

00

4

U
ni

te
d

A
irl

in
es

69

,2
65

,0
00

5

K
LM

22

,3
22

,0
00

5
N

or
th

w
es

t A
irl

in
es

55

,9
25

,0
00

6

E
as

yj
et

21

,9
17

,0
00

6
Lu

fth
an

sa

51
,2

13
,0

00

7
A

m
er

ic
an

A
irl

in
es

21

,2
28

,0
00

7
A

ir
F

ra
nc

e
49

,4
11

,0
00

8

S
in

ga
po

re
A

irl
in

es

18
,0

22
,0

00

8

A
ll

N
ip

po
n

A
irw

ay
s

49
,2

66
,0

00

9
E

m
ira

te
s

16
,7

48
,0

00

9

Ja
pa

n
A

irl
in

es

48
,9

11
,0

00

10

C
at

ha
y

P
ac

ifi
c

A
irw

ay
s

16
,6

67
,0

00

10

C

hi
na

S
ou

th
er

n
A

irl
in

es

48
,5

12
,0

00

S
ou

rc
e:

W
or

ld
A

ir
Tr

an
sp

or
t S

ta
tis

tic
s

(W
A

T
S

).
W

A
T

S
p

ro
vi

de
s

st
at

is
tic

s
fr

om
m

or
e

th
an

3
50

a
irl

in
es

, i
nc

lu
di

ng
lo

w
-c

os
t c

ar
rie

rs
. T

hi
s

re
pr

es
en

ts
a

c
om

pl
et

e
st

at
is

tic
al

p
ic

tu
re

o
f t

he
a

irl
in

e
in

du
st

ry

in
2

00
6.

E
xh

ib
it

3
So

ur
ce

: I
nt

er
na

tio
na

l A
ir

T
ra

ns
po

rt
A

ss
oc

ia
tio

n
(I

A
T

A
).

IA
T

A
, a

n
in

te
rn

at
io

na
l t

ra
de

b
od

y
cr

ea
te

d
m

or
e

th
an

s
ix

ty
y

ea
rs

a
go

b
y

a
gr

ou
p

of
a

irl
in

es
, r

ep
re

se
nt

ed
s

om
e

25
0

ai
rli

ne
s

co
m

pr
is

in
g

94

pe
rc

en
t o

f t
he

in
te

rn
at

io
na

l s
ch

ed
ul

ed
a

ir
tr

af
fic

.

0

20

40

60

80

100

120

140

20
00

20
01

20
02

20
03

20
04

20
05

20
06

20
07

20
08

20
09

20
10

Year

M
il

li
o

n
s

o
f

U
se

rs

Broadband Households
Total US Households

Exhibit 4: Frequent Business Travelers’ Use of Online Tools, 2002 and 200410
2002 2004
Total sample of frequent business travelers: 312 304

Number of respondents that:

Print boarding pass before airport arrival 30 77
Check in online 22 76
Receive mobile or PDA alerts 30 44
Use in-flight Internet access 7 7

Source: JupiterResearch

Exhibit 5: U.S. Broadband Household Projections, 2000–201011

Note: Mintel estimates are based on deployment data of all high-speed Internet access technologies, including cable service, DSL, and
both fixed and mobile wireless connections.

Needs help with similar assignment?

We are available 24x7 to deliver the best services and assignment ready within 3-8hours? Order a custom-written, plagiarism-free paper

Get Answer Over WhatsApp Order Paper Now

Do you have an upcoming essay or assignment due?

All of our assignments are originally produced, unique, and free of plagiarism.

If yes Order Paper Now