Analysing CRM At Air France
CRM Group Project, vendredi 14 novembre 2003
CRM at Air France Page 2 of 22
Table of Contents
1 INDUSTRY CONTEXT: AIRLINE CRM ........................................................................3
2 CRM AT AIR FRANCE......................................................................................................6
3 PROJECT REVIEW OF ISSUES, CHALLENGES AND RECOMMENDATIONS...9
3.1 ORGANISATIONAL POSITIONING OF CRM WITHIN AIR F RANCE ..........................................9
3.2 CRM WITH A “SERVICE ” FOCUS .....................................................................................10
3.3 SEGMENTING THE PASSENGER BASE .................................................................................13
APPENDIX: CRM AT AIR FRANCE ......................................................................................17
CRM at Air France Page 3 of 22
1 Industry Context: Airline CRM
The airline industry is synonymous with low returns. From 1990-2000 the average return on
invested capital over weighted average cost of capital (WACC) was 0.07%1 . Aggregated global airline
industry financial results declined steadily from 1997 to 2000, with operating margins falling from 5.6%
to 3.3% and net profit margins from 2.9% to 1.1% 2 . In 2001, aggregated operating losses stood at US$
10.9 billion and net losses at US$ 12 billion – on worldwide revenues of US$ 305 billion. Results were
equally dire in 2002. While industry analysts note signs of an up-tick in 2003, margins are expected to
remain exceptionally tight.
Such dismal results are a product of decreased load- factors (over-capacity), high labour costs, and
– catalysed by the events of 9/11/2001 – increased insurance premiums. In addition, full- service airlines
have had to grapple not only with the effects of the global economic slow-down but also with the
encroachment of low-cost carriers. Discount operators such as Ryanair enjoy compound annual growth
rates of up to 20%, while most other carriers languish around 4% 3 .
Unsurprisingly, airlines have been focusing much of their efforts upon improving their results
through cost reductions and efficiency gains at the operational level. However, increasingly airline
executives are also turning their attention to customer relationship management (CRM) as a means to
drive improvements to both top and bottom lines.
The precursor to airline CRM was the frequent flyer programme (FFP) – an initiative pioneered by
American Airlines in the 1980’s and subsequently adopted by most of the industry. The prime objective
of FFP’s was initially to increase customer loyalty. FFP’s have long been commoditized and are now
merely another cost of doing business. In the search for competitive advantage and elusive profits airlines
are now looking to transmogrify the way they manage their customer relationships.
Airlines face a number of significant challenges to evolving CRM in a meaningful, value-
Customer segmentation. Most airlines currently base their segmentation on miles flown as a
proxy for customer value. However, such segmentation is not only incomplete, but can also be
1 Compustat - Bloomberg
IATA World Air Transport Statistics, 2002
Hyped Hopes for Europe’s Low-Cost Airlines: McKinsey Quarterly 2002 #4
CRM at Air France Page 4 of 22
misleading. One major US airline recently discovered that its top 1000 customers by revenue had a 60%
higher impact on its bottom line than did its top 1000 customers by mileage 4 . For CRM to help guide
resource allocation and investment decisions, airlines need to identify a way to segment by “real”
customer value. They must also identify customer needs on a per-segment basis and gain a truer
understanding of how to service these needs in a differentiated fashion. In essence: along how many
service dimensions and customer segments can air travel genuinely be de-commoditized both within and
Technology and information fragmentation. Airlines interact with customers through multiple
touch-points; ranging from travel agents, call centers and internet portals to lounge, check- in and on-board
personnel. Most have also engaged marketing partners as adjuncts to their FFP initiatives. For example,
Citibank visa and Hertz car rental both tie into American Airlines’ Advantage programme. As a result, it
is not unusual for airlines to have customer data gathered across as many as 20 different internal systems
and 10 external ones 5 . Almost inevitably, system incompatibility issues and inconsistent data collection
standards mean that airlines’ experience of the customer is fragmented into vertical informational silos,
while the customer’s experience of the airline cuts horizontally across all touch-points. This incomplete
view of the customer can generate service shortfalls and dissatisfaction along with sub-optimal
deployment of resources.
Organizational design and management. By their nature, airlines tend to span multiple
geographies and cultures. This exacerbates the information fragmentation issues outlined above. In
addition, other important questions are raised. Presuming the informational problems are resolved at a
systems level, how can an airline instill a universally customer-centric mentality in its personnel? Also,
how can the airline provide its people with operationally viable guidelines concerning the way in which to
handle and differentiate between customer segments? Regardless of the industry involved, an estimated
55% of CRM projects fail to produce results6 and 25% actively damage customer relationships 7 . It is
estimated that 87% of such CRM failures are due not to technology or related factors, but to inadequate
The future of CRM in the airline industry. IBM Institute for Business Value 2002
CRM in the air: McKinsey Quarterly 2002 #3
Gartner Group statistic
CRM at Air France Page 5 of 22
change management and implementation at the personnel level8 . Getting the organizational side of CRM
right is thus crucial.
Southwest Airlines - perhaps largely as a result of its comparative youth, founding philosophy,
organizational model and the cultural homogeneity of its regional scope – has been successful in
embedding the CRM service mentality throughout its organization. This has been achieved through a
combination of employee training, service-centric incentives and operational empowerment facilitated by
the appropriate technological tools. The company’s core mission statement is “Dedication to the highest
quality of customer service delivered with a sense of warmth, friendliness, individual pride, and company
Other airlines have been less holistic in their approach to CRM – focusing instead on what they
believe to be individual differentiating elements. United Airlines has introduced its “Make Amends”
programme 10 - dedicated to addressing and resolving customer concerns pro-actively. Customers are
advised of cancellations and schedule changes prior to their arrival at the airport and anticipatory re-
booking arrangements are made. “Make Amends” activities are prioritized according to United’s
customer value segmentation.
Another carrier has concentrated upon leveraging its marketing partners’ information to gain a
more complete understanding (by inference) of the degree to which its existing customers share their
business with competing airlines. It does this by analyzing data from its co-branded credit-card –
identifying customers who are redeeming FFP miles through credit-card spending in hotels, restaurants
and shops in foreign locations to which they have not flown with the corresponding carrier. These
passengers then become the focus of targeted marketing by the airline.
For this project, we ha ve focused our attention on Air France – an airline currently wrestling with
the challenges of CRM, at both the strategic and implementation level. Air France has been kind enough
to provide us with access to both information and personnel related to its CRM initiative.
The future of CRM in the airline industry. IBM Institute for Business Value 2002
The future of CRM in the airline industry. IBM Institute for Business Value 2002
CRM at Air France Page 6 of 22
2 CRM at Air France
Air France carries 130,000 passengers on a daily basis. The company interacts with these
customers via 15 touch points 11 , the most significant of which are detailed below:
• Frontline personnel – including booking agents, check- in staff, lounge and on- flight staff. Air
France employs 25,000 such personnel. A new initiative handled by frontline operatives is
“ROC” (Rappel Operationnel des Clients), a proactive problem-solving and re-booking service
analogous to United’s “Make Amends” programme touched on in the previous section.
• 3rd party travel agents, both on- and off- line. Most travel agent booking activity interaction with
AF is managed via the Amadeus booking system.
• Air France’s website, providing online services such as travel-planning, scheduling and booking
– together with access to AF’s FFP, “Frequence Plus”.
• Opodo: a web-based “Pan European” travel company. It is a joint venture 12 (of which Air
France is a founding member) born mainly to compete against Orbitz.
• Call centers – offering customer service via both live personnel and interactive, automated
• Self-service ticketing and check- in booths at airports
• Air France’s marketing partners (eg, Hertz car rental and Accor Hotels) affiliated to the
Frequence Plus programme.
The back-end systems supporting each touch point have largely been developed independently of
one another and have yet to be fully integrated. Information is thus split into vertical silos - a problem,
which Air France recognizes and is working to address 13 . To compound this further, Air France segments
customers differently for each touch-point/information silo. This risks highly heterogeneous treatment of
the same customer across their points of contact with the airline.
See Appendix for full details
Opodo is owned by Aer Lingus, Air France, Alitalia, Austrian Airlines, British Airways, Finnair, KLM, Lufthansa &
See appendix for further details of Air France’s existing “Commercial Information System” framework
CRM at Air France Page 7 of 22
Perhaps the most important component of AF’s current customer- facing offering is “Frequence
Plus” – its frequent flyer programme. Frequence Plus has 5 million members - 300,000 of whom belong
to the elite clubs: Red & Blue. It also ties in with the “Sky Team”14 alliance FFP. Frequence Plus is
presently the main source of customer information used to drive targeted marketing campaigns.
To date, approximately €51m have been spent on Air France’s “Commercial Information
Systems”. A further €22.9m are specifically budgeted for the CRM project, through until the end of 2004.
A small team has been appointed by Air France top management to handle the CRM project
globally. The team has been mandated to select appropriate CRM technology, harmonize information
flows across existing legacy systems, align AF CRM efforts with those of the Sky Team alliance, lead
change management across the company and ensure the project as a whole provides tangible returns.
Quite a challenge!
Air France has defined the following broad goals for its CRM strategy15 :
• Instill a “culture of service” throughout the company.
• Standardize customer information and segmentation criteria across all systems and touch points.
• Simplify customer segmentation in order to avoid confusing frontline staff and to permit
operationally meaningful differentiation/customization of service offerings and relationship
nature on the basis of customer-value.
• Generate operational savings through service simplification and asset deployment optimization.
Conceptually, Air France has split CRM into two levels: “Decisional” and “Operational”. The
decisional level involves using techniques such as data mining to understand customer needs and
behavioural trends - then using that information to guide strategic decision- making. Operational CRM
involves providing frontline personnel with customer information that can usefully be acted upon (rather
than a deluge of data that cannot be “operationalized”).
Sky Team members: Aero Mexico, Air France, Alitalia, CSA, Delta, Korean Air
Information obtained from interviews with Air France personnel, plus Air France documents (see appendix)
CRM at Air France Page 8 of 22
Interestingly, Air France’s CRM strategy appears to be focused on the service component of
CRM – while sales and direct marketing seem not to fall directly beneath the CRM umbrella. As for many
airlines, determining how to estimate customer value in order to drive segmentation is proving a challenge
for AF. Equally problematic is the calculation of a convincing ROI on the CRM project. These issues will
be examined in greater depth in the following section.
CRM at Air France Page 9 of 22
3 Project Review of Issues, Challenges and Recommendations
3.1 Organisational Positioning of CRM within Air France
AF’s organisational positioning of its new CRM department provides an insight into how its CRM
capability might develop. On the positive side, it has been given “direct report” status to the Chief
Executive’s Office – a clear sign that senior management wants to get the message across that the topic
needs to be taken seriously. The Head of CRM also has a reputation for “getting things done”. He was
successful in imposing a new operational performance measurement system across the airline’s
From a less optimistic perspective, the CRM team is very small (about ten people). This may be a
temporary phase (with the team taking on an “evangelising” role for CRM thinking within the airline).
We also got the impression from our meeting with the CRM Head that the team emerged only after a
significant investment had been completed in a CRM system. This implies a “we have bought the
technology, now we must build a strategy” approach. This strategy has been given a huge breadth of
focus; cited objectives include a differentiation of AF from the low cost carriers, an integration of AF
CRM into Sky Team alliance efforts, a recognition of corporate account revenue and even an alignment
with the company’s centralised purchasing policy.
However, the team is only beginning to think about how to define its relationship with the other
functional areas within the airline. For example, should it act as a coordinating body between Information
Management, Sales and Distribution, Loyalty Management and Operations? Or should it reside within
one of the core departments and use this leverage to drive policy across the airline? Should it focus on a
narrow initial plan and seek to win over converts with a demonstration of success in a limited area (i.e. a
more sophisticated use of existing direct marketing campaigns via CRM profiling)? Or should it try to
integrate different functional departments within an overall strategy in which all areas of the airline can
begin to grasp the power of focused customer relationship management?
Air France clearly needs to prove the CRM principle to its own constituents. The team needs to
build credibility in order to garner support for the project within the organization – as evidenced by their
request to us to find “real examples” of CRM with tangible ROI metrics. From our perspective this
means that they should pick a narrow area and trial it – ideally somewhere where it is easier to track “cost
and revenue” performance. We suspect that this would be hard to achieve in the “service” CRM
CRM at Air France Page 10 of 22
dimension as the data is too challengeable and the process will take too long to get up-and-running.
Instead we believe it would make more sense to look at one of the direct marketing campaigns and
understand how/if the CRM software that they already have (plus some of the CRM “cube” thinking that
we will examine in greater depth later in this document) could be leveraged effectively.
We also think that – in the long-term - the CRM team should be located within the AF Sales and
Distribution/Sales or Marketing function. CRM is supposed to be a company-wide philosophy, touching
all staff and functional roles. However, for AF, it needs to be part of a core, strategic, major revenue-
generating department. Information Management and Loyalty are too often seen as support departments;
while Operations/Customer Service are “cost” departments that generally lack strategic feel (or
3.2 CRM With A “Service” Focus
The Air France CRM programme has been initially positioned with a “service” focus. For
example, the CRM outline document (see Appendix) refers to the need for a “homogenous” service as a
key goal. At first inspection, this seems an unusual positioning given that the majority of CRM initiatives
aim to segment customers in order that they can be treated differently. However, discussions with AF
employees suggest that the “homogenous” system is really more of an attempt to recognise “high- value”
customers across various contact points within the customer experience (booking a ticket, checking- in,
lounge and after-sales). A differentiated service would then be offered.
Our concern with this approach has been that it seems to be too rooted in the “elite” end of AF’s
existing passenger base. CRM, at this early stage of the AF approach, is primarily a “notifying” system to
inform an AF employee that the person on the telephone or in the check-in queue should be handled with
more “service” than the average passenger. In part, this reflects that information is richest within the
tiered data in the airline’s FFP/Loyalty Management systems; it is natural that this is where AF might
want to begin its work. Clearly too it is beneficial to AF to attract and retain influential passengers (as
well as the top tier of passenger type that travels very frequently).
However, by aiming at the “top of the pyramid” AF may miss out on cultivating relationships with
large numbers of passengers lower down in terms of revenue performance and “status”. This is
particularly the case in failing to identify fut ure revenue streams from passengers who may have the
potential to generate significantly more AF income if they are managed correctly. It also risks ignoring
passengers who have yet to fly with AF (or who do not figure in their FFP databases).
CRM at Air France Page 11 of 22
Another observation on the “service” approach is that, although AF was clear in wanting to
understand how to measure ROI against its CRM initiatives, it also seems to define CRM performance
primarily in terms of its impact on “customer satisfaction”. The assumptio n is that “satisfaction” (driven
by recognition and elite handling of the “high- value” passengers during the AF passenger service
experience) will promote “loyalty” which will in turn generate “profit”.
Our assessment here is that it is notoriously difficult to measure a profitability (or even revenue)
return on broad notions of “satisfaction” and “loyalty”. Linkages between loyalty and “word of mouth”
brand equity, as well as between loyalty and “lower cost to serve” are not clearly proven. AF admits this
in its own strategic outline but suggests that “on a scale from one to five, an increase of 0.1 in customer
satisfaction currently represents 40M€ additional revenues”. We have not been able to identify how this
calculation has been made.
Finally, from an AF perspective, it is not clear that the concept of “satisfaction” criteria has been
thought through beyond an understanding of a “corporate-wide” satisfaction measurement process. As far
as their current understanding is concerned, much of the “service” focus is assigned to disruption
situations (calling the passenger in advance to warn him/her of a delay) or failed service scenarios
(handling for lost baggage or other service errors). Whilst this is laudable from a service recovery
scenario, is this sufficient focus for a CRM campaign? Indeed, should it even be the initial focus? Is
CRM being seen too specifically as an “operational enabler” rather than as a strategic tool?
Firstly, AF would need to decide how many different “types” of service that it wanted to offer.
Currently, in service terms, it segments its passengers according to the three classes of travel (with some
additional filtering according to FFP or corporate card status). If it wishes to differentiate by more than
this “three-plus” approach, it needs to be very clear how passengers would recognize the service-style to
be different. Would this mean a shorter queuing time at a call center? Would it mean a specific area of
the lounge is made available? Would AF provide an accelerated check- in process or a broader (any)
smile from the check- in staff? This may seem an obvious point but, before any further work is done on
creating distinct service styles via a CRM philosophy and technology base, AF must decide if its systems
(and especially its staff) will be able to meet the demands of the new approach.
Secondly, if AF did want to proceed with a “service satisfaction driven” CRM campaign it would
need to collect “satisfaction” criteria for each of the passenger segment types that it was seeking to
address. The criteria would need to be disaggregated and weighted according to “satisfaction” drivers of
CRM at Air France Page 12 of 22
the different segments. The service activities most relevant to achieving high scores in each of these
criteria would then need to be identified. The customer “touch points” at which the “satisfaction” data
would be collected, analysed and acted upon would need to be mapped out.
In short, a “service focused” CRM would be a massive undertaking if it were to be focused on
targeting “high- value” customers, identifying how to increase their “satisfaction” with AF, measuring
performance and linking all of this back to revenue impact. As we have our doubts as to the reliability of
the “revenue” or “profitability” measurement process (as well as a keen understanding that the CRM team
needs to demonstrate practical achievements in a near-term time frame), we wonder if this effort would be
less fruitful than a more targeted “sales CRM” approach.
From discussions with another airline it was clear that the difficulties in linking “satisfaction” to
revenue also exist outside the context of Air France. The other airline commented that it was far easier to
judge ROI on targeted sales campaigns than on broader “service” efforts. The airline had developed a
test-and-control model for defining a target audience (against revenue, geographical and behavioural
data). It then tested an offer within a sample population, assessed ROI on the CRM campaign relative to
the sample and then implemented the offer.
Our interview also indicated that as a result of narrower campaign successes the “CRM capability”
was beginning to move in- house within the Revenue Management department of the airline. Route
controllers were identifying specific sectors or time periods when load factors were poor and a suitable
sales offering was established via liaison with country sales departments. Involvement with CRM as a
strategic tool was growing, often building on practice that already existed within the airline (i.e.
experience of direct mail campaigns, understanding of frequent flyer travel patterns, assessment of
competitors’ products etc). In this way, airline personnel began to understand CRM “from the ground up”
and made explicit linkages between what might be possible in the future through a more sophisticated
CRM strategy from the building blocks of their existing experience. It has been an iterative approach,
with CRM diffusing slowly within the airline. Collection and interpretation of raw data – cost of
campaign, take-up, revenue and “rough” ROI calculations – was helping in this process.
We reiterate that AF should explicitly recognise that both “targeted sales” and “service”
dimensions fall under the CRM umbrella – focusing first on targeted sales to drive measurable ROI.
Strategically, we suspect that this will be key to the sustainability of the project as a whole. The initial
CRM at Air France Page 13 of 22
excitement of top management concerning CRM may well subside if large investments are not seen to
have a measurable return. CRM thinking is still in its infancy and algorithms convincingly tying customer
satisfaction to loyalty, and loyalty to profitability, virtually non-existent. The CRM team at Air France
needs to buy time in order for CRM thinking to evolve and for them to lay the foundations for subsequent
service- focused initiatives.
Regardless of the service-driven measures that Air France subsequently decides to implement, it
needs to address two issues. The first is technology based and involves welding together information
currently fragmented across disparate systems and customer touch-points. The second involves forming a
clear picture of how to measure customer value and segment the customer base accordingly. How can AF
avoid jumping into “operational” CRM before it has identified more clearly how it will support
operational performance through its “decisional” CRM strategy? This second issue is examined in greater
3.3 Segmenting the Passenger Base
An effective segmentation strategy underlies both the “sales” and the “service” approaches and
recognises “higher-value” customers in both of these areas. AF was very open in admitting that it does
not have an accurate picture of which of its customer segments are “higher value” and it asked us to help
in working this out.
Firstly, we asked who AF’s most profitable customers are today. However, in discussion with AF
and other airlines we discovered that this was a difficult question to answer. Reliable “customer
profitability” calculations for airline passengers do not yet exist. Personalised or even aggregated cost
information is too impractical to collect and either “mileage” or “revenue” data is used instead as a proxy
for profitability assessments.
Secondly, as AF does not have a clear idea of how to manage the “lifetime customer value” notion
of its diverse passenger base, we asked who we should expect to be more profitable customers in the
future. Other airlines are more advanced in this area and have begun to consider this issue (primarily by
using loose behavioural and attitudinal assessments alongside existing data on passenger revenue, route
and frequency performance). FFP retention rates offer an opportunity to calculate potential lifetime
durations. If a generic value of annualised passenger trave l is then applied to the lifetime duration and
discounted back, AF can build up an estimate of future customer value (see an example format in the
“cube model” construct).
CRM at Air France Page 14 of 22
A three dimensional segmentation model
Despite the increased sophistication of the CRM tools available today, as well as the more
sophisticated thinking that is beginning to emerge within many companies on how to leverage this
technology, CRM strategy still seems to be evolving through an iterative, trial-and-error process 16 . How
should AF begin this process when the opportunities and challenges facing its CRM team seem so
Below we have tried to conceptualize how AF should develop its basic thinking on the
foundations of its CRM strategy. We recommend that AF create a segmentation principle that on one
hand builds upon current business practices and information available, while on the other hand it takes a
step towards key CRM principles like customer lifetime value. The segmentation scheme we propose for
AF i built around three factors: Past value, Future value and
Tactical Value (VIP, Corporation)
Tactical value. Past value is important because it rewards
loyalty. Future value is of course important because by acting on
it AF helps drive retention of valuable customers. By Blue
incorporating a factor for tactical value AF can overlay the
previous segmentation with a qualitative assessment of which
customers and corporations are important. The need to capture Future Value (Estimated)
this “VIP factor” was something that was pointed out to us by
AF. Furthermore we recommend that each factor be segmented Past Value (Frequent flyer miles)
according to AF’s current scheme Basic-Blue-Red.
One advantage of this approach is that the building blocks for this segmentation already exist
within the FFP programme. This will mean that they are biased towards existing passengers. However,
this would still provide an important learning process for AF and we believe that the segmentation that
they should develop will also reveal insights to the types of “potential” passengers that should also be
targeted. In that sense the thinking will go from being descriptive to prescriptive.
At Capital One, hallowed CRM pioneer, science is mixed with perpetual experimentation. Interviews with a Capital One
employee at INSEAD pointed to the constant modifications of data, hypotheses and trials going on. Capital One benefits from
more sophisticated data management and interrogation technologies than most but the organizational philosophy to experiment
is a key driver of its success.
CRM at Air France Page 15 of 22
Assessing the value of a customer
It is difficult for us to define exactly which data points should be used in the segmentation
calculations. Deriving such parameters lies beyond the scope of this project. Nonetheless we feel
comfortable pointing to some areas for consideration:
Past value. Given that past profitability information is not readily available at AF, we believe that
AF can use the standardised approach in utilising flown revenue data from its existing FFP programme as
a proxy for past profitability.
Future value. AF must begin to think about how to create basic models, based on past data, that
group roughly similar passenger types and profiles (for example, “small business Europe” or “biannual
full fare international leisure”) into future revenue streams. Some work has already been done on the
attitudinal characteristics of five core segments but AF needs to attach fare-type, purchasing behaviour,
travel behaviour and revenue estimations to these groupings. This analysis should also give AF a good
estimation of the average retention rate per group.
The estimated lifetime value of a customer can be derived with a simple discounted cash flow of
estimated gross operating margins. In this context we believe that it is possible to use gross margin or
“revenue contribution” (an industry standard for measuring revenue contribution per sector flown). But, a
key to a successful CRM adaptation of the formula below is to get a good handle on projected retention
rate. Here the rate derived from past data should be supplemented with individual demographic data such
as age and residence location. For example, the customer group “Business traveller France” may have a
high overall retention rate of 0.9. However, an individual business traveller who has just turned 67 should
have a lower estimated retention rate. Consequently, that person will have a lower projected future
revenue (but may of course still be important due to past or tactical value). In summary the formula for
calculating future value is as follows:
Gross Margin t ⋅ r t ⋅ f (demographi cs )
Future Value = ∑
t =1 (1 + d ) t
where r = retention rate from 0-1 and d = discount rate
Tactical value. A higher degree of tactical value can be attributed to individuals, groups of
individuals (such as the grouping from above), corporations and groups of corporations. As an example of
the latter, given the upcoming AF-KLM merger, one can imagine a situation where it becomes of strategic
CRM at Air France Page 16 of 22
importance to AF to keep Dutch corporate clients. In this case AF could assign these passengers a high
Applying the cube segmentation Tactical value
The resulting cube based on this 3-axis segmentation
fulfils the dual purpose of being insightful yet simple to apply.
Depending on the application / area of focus the cube can be
sliced and grouped in multiple ways (see examples below).
However, unlike previous segmentation schemes at AF, all
segmentation would now be based upon a common model.
Service segmentation could be done as follows.
Anyone scoring red on any one factor would be entitled to red (premium) service. Similarly, anyone
scoring blue on any one factor is entitled to the blue service level. The rest would be entitled to basic
service. In this way there are still three (3) simple classes, in line with what AF service personnel is used
to, yet based on a more insightful segmentation. The cube also offers the possibility for further
segmentation - for instance passenger prioritization for class upgrades. Passengers scoring high on more
than one axis could be given the status “deep red”, efficiently indicating their importance to AF. See
Basic service Blue service Red service
Tactical value Tactical value Tactical value
Future value Future value Future value
Past value Past value Past value
For sales segmentation, the dicing of the cube could be more diverse. It could be sliced more
finely for targeted sales campaigns, as sales offerings can be more easily personalized in practical (route,
date, time, price etc) than can specific forms of operational “service”.
Appendix: CRM at Air France
(Translated from internal Air France document)
For many years, Air France has spent a lot of money to build its loyalty program. Furthermore, thanks to
important investments in its information systems over the past few years, the company currently owns a
sufficient amount of data to develop an effective CRM system.
Personalisation of the relationship with customers through the chain of service is part of the company’s
global strategy for the years to come. CRM is the pilar of this strategy.
2. Current situation in terms of client relationship at Air France.
(a) Many efforts have been made to improve this relationship recently:
Over the last 3 years, an active loyalty policy has been developed around Air France frequent flyer
program – Frequents Plus. It has today 5 million members of which 0.3 million are in the Clubs
(Red & Blue). Beside Frequents Plus, Air France has launched many initiatives all along the chain
(i) Reservations: Internet, interactive vocal system, reorganization of the call centers.
(ii) On ground and In-flight: Redefinition of the products and services; plan of action at the
CDG hub, “ROC” (“Rappel Operationnel des Clients”, automatic rebook of passengers
prior to their trip to the airport if their flight is cancelled).
(iii) After Sales: Improvement and simplification of procedures, especially for luggage issues
(introduction of a dedicated call center).
(iv) Direct Marketing: Implementation of a sophisticated direct marketing tool allowing perfectly
Furthermore, numerous horizontal projects have been launched in order to facilitate the interaction
between the customer and AF: Implementation of a self-service chain (Internet, e-ticket, self
checking machines at the airport), launch of an ambitious service education program for the front
line agents, service certification, homogenisation of the service worldwide, and development of a
In total, AF spent 51 million euros in its commercial information system so far, and is planning to
spend an extra 22.9 million euros until the end of 2004. Overall, the relationship with the customers
represents a budget of approximately 2 billion euros.
(b) However, there a consensus emerged within the company to identify some weaknesses.
(i) First, the quality of the relationship with the client and its level of personalisation are
discontinuous and heterogeneous from one point of contact to the other.
(ii) Second, the fact that there are different segmentations of the customer for different
functions of the company creates confusion in the customer’s treatment by the frontline.
From the client’s point of view, this creates heterogeneity in treatment.
(iii) The relationship with the client is historically focused on passengers, but not on corporate
accounts. However, corporate travel representing most of AF’s profits, it is necessary to
take this dimention into account.
(iv) Real improvements can be made in the after sale aspect of the relationship with the client,
particularly for those who have been involved in operational problems (delayed or
cancelled flights, lost luggage…)
(v) At this stage, the company doesn’t have a consolidated economic vision of CRM (how
much does it cost ? what is the value of a client ? how much to invest on him ?...).
(vi) All the information already available (Frequence Plus, marketing database,
datawharehouse…) are underutilized or not utilized at all by the front line agents,
considerably limiting the return on this important investment in the commercial information
(vii) There is a lack of definition of the role of each point of contact in CRM as well as an
absence of defined interactions between the different points of contact.
(viii) The important turnover of employees in contact with customers creates growing
heterogeneity between new and older agents.
(c) Furthermore, AF CRM project must integrate the following evolutions:
(i) How to differentiate the Air France offer in the context of the low-cost airlines offensive?
(ii) How to integrate the fact that through the Sky Team alliance almost half of the Air France
flights are not done with Air France planes?
(iii) How to answer the clients’ expectations that are increasingly demanding, particularly in
terms of recognition and personalisation?
(iv) How to take into consideration the restructuring, distribution and the centralisation of the
company’s purchasing policy?
(v) How do you value such a project at a time when the company is heading towards the
(d) Therefore, the Air France CRM project wants to achieve the following:
(i) Treat the client in the continuity of his or her travel with seamless treatment throughout all
the chain of service.
(ii) To help the actors of the frontline by redefining their roles and objectives and giving them
adapted tools, especially in terms of their work environment.
(iii) Create the conditions for developing an homogeneous culture of service.
(iv) Define a financial strategy where the resources of the company are wisely allocated, taking
into account the value of each customer.
3. AF CRM: Definition of the goals
(a) Project definition: satisfaction of the “travel contract” for all the customers and personalisation for
The management of the company has identified three levels of customer relationship to take into
A. A personalised level that has to take into account the long term relationship with
regular clients or high potential clients. This personalisation will be more efficient if
the clients can be distributed in homogeneous segments.
B. A level of relationship that takes into account certain “communities” of clients:
Corporate accounts, passengers with certain particularities…
C. A last, a “basic” level corresponding to the ”travel contract” specific to each class of
service: First, Business, and Economy class.
Gestion personnalisée de la
Personnalisation : vraie valeur ajoutée perçue relation client
One to one
Considération/reconnaissance : valorisation perçue Gestion différenciée de la
One to few relation client
Choix : ouverture et respect
Innovation : modernité et surprise
One to many
Facilité : l’avion facile
Séduction (réponse aux cinq sens) : le charme
Respect du contrat de base : fiabilité et transparence Gestion de la relation client
One to all
Sécurité : le préalable
For AF, the CRM project has a double purpose:
(ii) Continuity and homogeneity throughout all the chain of service.
(iii) Identification of the points of contact where personalisation is possible and generates ROI.
(b) The two conditions of success: Involvement of the employees and easy access to commercial data.
(i) Involvement of employees.
One of the main conditions of success for the CRM project within the company is the participation of all the
categories of employees in contact with the clients: call centers agents, airport agents (checking, lounges…),
flight attendants, pilots, sales representatives, after sales agents, … the challenge for AF is to create a
common culture, a same client segmentation, and common tools across all these functions. At some point,
this will probably involve the creation of bridges for employees between these points of contact. It also
implies their participation at all the stages of the project implementation.
(ii) Capacity of the company to handle information
The two challenges are the improvement of information gathering and a better communication of the
information between the different points of contact of the company. In fact, despite the tremendous amount of
information currently available, each point of contact has its own vertical interactive software dedicated to
handle relationships with the customers. The graph below gives a general idea of the existing system:
Avant Voyage Sol Vol SAV Fidélisation
Centre Freq +
(A) (F) (Public) (Public) ( Aff.) (Public)
Web B2C (¨AF.com )
Borne Libre Service
Web B2C (xx .com )
Interaction Smash PIL
AMADEUS (Résa) ROC
Transaction Ticketing AF
Ticketing A F World
Serveur Client Tracer
Référentiels Référentiel Firmes // Référentiel Agences // Référentiel Individus + BDM
Référentiel Firmes Référentiel Agences Référentiel Individus + BDM
Décisionnel DWH Réseau / /DWH Client
DWH Réseau DWH Client
Application Air France Papier
Furthermore, the information update process, as it is handled today, is a real obstacle to an homogeneous
database. AF is conscious that integrating all these separated information systems into a common, user friendly
base is a prerequisite for any further investment. The other issue concerns the level of detail necessary to front-line
4. AF CRM: Implementation
A dedicated CRM team has been appointed by the Air France general management a few months ago. This
team has a double mission:
• Pilot, consolidate, and insure the coherence of all CRM initiatives in the different functions having an
impact on customer relationship.
• Launch and lead horizontal CRM initiatives that no function can manage on its own.
(a) How projects are prioritized
Several criteria are used to prioritize CRM projects within the company :
• Breach of coherence of customer handling in the service chain
• Points of contact that are ideal for service personalization
• Projects with higher potential economic gains
• Upgrade of service offer when necessary for the respect of the “basic” travel contract.
• Potential for employee participation
(b) The CRM team activities
• CRM project management from a global perspective
• Economic monitoring of the project
• Change management within the company
• Communication within the company
• Centralization of information systems needs
• Coherence with AF’s general marketing policy
• Coordination of current projects
• Harmonization of CRM within the Skyteam Alliance and ith the domestic partners
(c) The CRM team main projects
The company made a distinction between “horizontal” and “functional” projects. Horizontal projects
are concerning the entire (or most of) the service chain. As a first step, Air France is going to focus
on 4 horizontal projects:
• Developping an homogeneous culture of service within the company
• Integrate and share all the relevant data all along the service chain.
• Qualify and keep historical customer data
• Keep the economical control of CRM
Functional projects are focusing on specific points of contact of the service chain : Sales (in
particular corporate sales, internet), on ground (lounges, airport services…), in-flight, and after
sales (lost luggage, customer reclamations…). Their objective is to identify client needs and
implement solutions in terms of tools, processes, competences, and attitudes.
5. Estimate of economic gains.
CRM economic gains are difficult to value. However, an estimate of the following potential returns is being
• Direct additional revenues generated by the CRM implementation
• Productivity gains linked to increased efficiency
• Direct savings linked to implementation of new CRM tools. eg: ROC project (“Rappel Operationnel
des Clients”), automatic rebook of passengers prior to their trip to the airport if their flight is
• Qualitative gains linked to customer satisfaction increase and consequently, the will to fly again on
AF. The company is conscious of the difficulty to value these gains but it made the following
estimate: On a scale from 1 to five, an increase of 0.1 in customer satisfaction currently represents
40M€ additional revenues