Nordic Insurance Company performance - The customer view on quality and value

Artikkelforfatter: Jan Eklöf
E-mail: jan.eklof@epsi-rating.com
About:

Jan Eklöf

 

is Associate Professor at Stockholm School of Economics and CEO of EPSI Research Services. His research is focussed on quality management and performance assessment of commercial as well as public sector activities.

 

 

 

 


Udgave:
2, 2008
Språk: Engelsk
Kategori:

 

Introduction

In the present study we illustrate and analyze how systematic measurement of custom-

Any organization has to listen to its external er satisfaction, loyalty and its driving forcescustomers and other stakeholders. A number can be used to highlight similarities and difof studies have shown that the long-term ferences in the Nordic insurance markets.

success of a corporation is closely related to This in turn gives indication about the funcits ability to adapt to customer needs and tioning of the various markets and specificchanging preferences. Customer Satisfaction

is a crucial goal for most organizations, and it Jan Eklöf is Associate Professor at Stockholm School has proven effective to forecast the long-term of Economics and CEO of EPSI Research Services. His research is focussed on quality management and

financial performance of any company (Itt

performance assessment of commercial as well asner, Larker, 1998). public sector activities.

relationship between suppliers and customers. It is further shown how such information may support improvement work and company specific priority setting and evaluation of strategies.

The next section gives a basic foundation for the used methodology, while some Pan Nordic benchmarks are given thereafter. A few results on causal relationships and usage of the model approach for analysis are given in the latter part of the article. It is ended with some proposals and ideas for corporate analysis and references to further reading and related studies.

Analysis of Customer Satisfaction in a Causal Model

What is classified as a good, acceptable or appropriate level of satisfaction, retention or customer loyalty? This depends on how it is measured and presented, but even more so, on the expectations and targets set for the activity. The concept of benchmarking is increasing in use for relating one actor's performance to another's in a systematic way and for learning how to strive for improvement. Benchmarking the framework for comparison has to be duly calibrated and studied sets of data have to be "harmonized". The following characteristics of a system for EPSI at national level are adhered to:

Constitutive Characteristics of EPSI at national level1

(1)

a measurement of satisfaction as well as key success factors

(2)

of companies and institutions

(3)

comprising several industries/sectors of

(4)

a nation or a distinct economic area

(5)

via periodical analyses

(6)

by a neutral institution.

For measuring Customer Satisfaction in a wider context a structural equation model has been developed based on components considered crucial for causal analysis (cf. EPSI Rating 2007). Thus, the model contains the components:

Image

Customer Expectations

Customer Perceived Product Quality

Customer Perceived Service Quality

Customer Perceived Value

Customer Satisfaction

Customer Loyalty The components are viewed as latent variables determined by a set of manifest constructs for each component. Each component is measured, the level of each component estimated, the relevant connections between the components established and the magnitude of the connections estimated. The objective is to provide results that are relevant, reliable, valid and have predictive financial capability. The structure of the model is illustrated in Figure 1. In order to estimate the model, to calculate the index values and the magnitude of the impacts, data must exist that is representative for the customers of the company/institution

in the country. The survey design considered consists of:

The design of a questionnaire for measuring the latent variables via the manifests. The questionnaire contains the manifest variables and some background variables.

The selection of a representative sample of customers.

The field operations for collecting the data from the sample using the questionnaire. This also includes the consistency checking and the creation of data files for analysis.

The analysis of the data and the preparation

of reports. A set of master questionnaires has been developed and their performance tested. The questions related in blocks to the latent variables. Each latent variable was covered by at least 3 questions concerning different aspects of the

Figure 1: The EPSI Rating Structural model

 

latent variable. The respondents were requested to respond on a scale from 1 to 10 where 1 means the customer was not at all satisfied with the specific aspect under study while a 10 means the customer is completely satisfied. In all around 30 to 35 questions must be answered to fulfil the requirements of the model. Three specific questions are related to the Customer Satisfaction variable. They cover:

The overall satisfaction with the product

Satisfaction compared to expectations

Comparison to an ideal product For the Loyalty latent the questions cover:

Intention to buy again

Intention to recommend the product

Price sensitivity assessment (in a few coun

tries) On the driver side some 15 –25 questions are used depending on the characteristics of the industries under study, but the wording of these questions is to a great extent standardized.

The model is estimated using Partial Least Squares (PLS) techniques. The 1–10 scale is, for ease of interpretation, transformed into a 0 –100 scale. The values of the impacts were estimated together with error margins. A separate model is estimated for each company where reliable results can be obtained. For the sector the average indexes were calculated using weights derived from the market shares of the companies. The data could be further disaggregated according to the background variables to study some sociological properties.

Each estimated model is empirically evaluated and clear-cut quality assurance indicators are given. Such Quality assessment measures are monitored for all steps of the process. These are related to the following aspects:

Validity (relevance aspects);

Availability;

Simplicity;

Comparability.

From a statistical perspective the models must possess enough predictive power and an ability to explain variation in data to make it meaningful for drawing inference. From a more general stakeholder perspective the models developed must be easy to understand and be relevant for the various users. The presented estimated models are empirically evaluated regarding:

Reliability (precision, coverage and other accuracy components);

Robustness;

Predictive Power. All these aspects are important in their own right. At the same time it should be observed that they are highly inter-related and require simultaneous attention in the design stage. Thus, fulfilment of acceptable levels for these criteria calls for strict handling of data collection throughout the process. The following minimum quality requirements are agreed upon for each individual model to be part of the EPSI Rating approach.

Any substantial failure to fulfil any of these must be commented and the reasons analysed.

Precision

The 95 percent confidence interval for the customer satisfaction index must have a width of not more than 4 units (referring to an index scale of 0–100)2 .

Explanatory power

The econometric model must be able to explain approximately 65% of what drives customer satisfaction.

On average an error margin of 2 units has been realized for the indices (measured by a 95 percent confidence interval). This may be interpreted as confidence interval with 95 percent probability covering the true value (lies between the estimated value plus/minus 2 units).

Satisfaction with the Insurance company – National benchmarks

The Insurance sector has been measured in the majority of Nordic countries since year 1999 in annual studies. The coverage has successively been increased and in 2007 both General Insurance (final Consumers and Corporate Customers) and Life/pension insurance (final Consumers) are covered in the majority of the Nordic markets. Below a few key trends and benchmark results are given.

General Insurance Private Consumer segment

In the diagram below, the private consumers in the respective country give the scores of their General Insurance Company. The average for the entire region (calculated as the GDP-weighted average of the respective country result) has increased significantly since last year.

The Icelandic consumers give the lowest average scores, while Sweden falls on second place from the bottom. Denmark and Finland are on top. The highest relative increases are reported for Finland and Denmark during the entire period studied, while Norway has been stable sine 2005.

Table 1. EPSI, General Insurance Private Consumer segment, Nordic countries, 2004 - 2007

Final Consumers Satisfaction

Country 2004 2005 2006 2007

Denmark 73.0 74.2 74.6 77.7 Finland 70.7 70.6 75.4 Iceland 69.7 68.1 66.7 Norway 64.6 69.6 68.6 68.9 Sweden 65.9 66.7 67.2 67.6

The Nordic Countries 67.6 69.9 69.9 71.3

The satisfaction has improved significantly in both the Danish and Finnish sectors during the last year, while a small increase is registered for Sweden, a decrease in Iceland and fluctu

Graph 1. EPSI, General Insurance, Nordic countries, 2004-2007

General Insurance EPSI Nordic Satisfaction 2004 - 2007

Satisfaction

80

75

70

65

60

ations in Norway. The difference between the highest country average year 2007 (Denmark) and lowest (Iceland) is 11.2 units. The number of companies and spread in satisfaction within each country differ also quite a bit. This gap is taken as an indicator of the actual competitive situation in the respective nation The highest score for the Nordic region is obtained for the group "other insurance companies" in Denmark (with 83.4) and the lowest score in satisfaction for one of the main companies in Sweden (63.3).

The scores for insurance can be compared to the average for the economy in the respective country in order to tell whether this sector is strong or weak in terms of market transparency. This is illustrated in table 2. It is seen that the insurance sector is higher than the national average in Denmark, Finland and Norway, while on par with the national average in Iceland and lagging slightly behind in Sweden. For the Nordic region as such the Insurance sector obtains higher scores than the entire economy measured.

 

Table 2: EPSI Economy scores compared to General Insurance 2007by country

Country Economy Insurance

Denmark 70.3 77.7

Finland 72.6 75.4

Iceland 66.5 66.7

Norway 66.4 68.9

Sweden 68.4 67.6

The Nordic

Countries 68.8 71.3

The profile of general insurance sector according to EPSI is illustrated in graph 2. The scores here relate to the model aspects above from company image through the other drivers to satisfaction (CSI) and perceived Loyalty.

Graph 2. EPSI, Profile General Insurance, B2C, Nordic countries, 2007

Profile General Insurance B2C Nordic EPSI 2007

85

80

75

70

65

60

It is seen that the basic profile coincide between countries. At the same time a few special features are observed, like:

Image is highest in Finland, and lowest in Sweden;

The gap between expectations and perceived quality is largest in Norway and Sweden;

The perception of value (for insurance fees) varies largely with Denmark and Finland on top and Sweden lowest according to the final consumers; and

Perceived loyalty is lower than satisfaction in Denmark, while the opposite holds for Sweden and Iceland, illustrating a strong switching barrier in these countries.

General Insurance corporate customer market

The coverage of also corporate Markets for General Insurance has successively been increased. The results for the last few years are illustrated in Table 3 and Graph 3.

 

Index-level

Table 3: EPSI - General Insurance, Corporate Customer segment, Nordic countries, 2004- 2007

Corporate customers Satisfaction

Country 2005 2006 2007

Denmark 70.6 73.4

Finland 72.6 74.2 73.3

Norway 67.1 65.5 68.5

Sweden 65.3 66.9 68.8

The Nordic

Countries 67.6 68.9 71.0

This means that the satisfaction among professional (corporate) customers (B2B segment) is almost as high as for final (individual) consumers (B2C segment). It has also improved more for corporate customers on average than for final consumers.

Graph 3. Corporate Customer satisfaction EPSI 2005 - 2007

General Insurance Corporate customers EPSI Satisfaction 2005 -2007

Satisfaction

80

75

70

65

60

Life/Pension Insurance, brokers and vehicle insurance in Sweden

Above, some core results for the general insurance industry were given. Within the insurance sector also other areas are successively being included in the regular EPSI studies. At present we have life insurance covered to some extent, as well as emerging studies for the insurance broker industry and special for the motor vehicle segment.

Time trends for the life/pension insurance segment are only available for Sweden so far, as this segment only recently was introduced into the studies in the other Nordic countries.

The Swedish life insurance market and the individual companies are illustrated below.

Denmark Finland Norway Sweden The Nordic

Countries

Country

As can be seen Skandia constitutes a special case, and still exhibit a satisfaction level more than 10 units below any of the other measured companies. From the more detailed analysis it is clear that main explanatory factors for the low satisfaction with Skandia still has to do with image and service quality, while the score for product quality is (almost) on par with the industry average.

The main driver for explaining the satisfaction with the life insurance company, apart from the performance of the savings, is trust in the corporation. The explanation of the insurance product(s) that a life insurance customer has needs also further improvement according to the respondents to the survey.

Satisfaction level

Graph 4. Customer Satisfaction Swedish Life insurance companies

Customer Satisfaction Life/pension Insurance Sweden Final conumer SKI 2001 - 2007

65

60

55

50

45

40

 

The main driver for explaining the satisfac-has needs also further improvement accordtion with the life insurance company, apart ing to the respondents to the survey.from the performance of the savings, is trust in In parallel to the above sketch, the rating bythe corporation. The explanation of the insur-the corporate customers for life-insurance/ance product(s) that a life insurance customer pension are as follows.

Graph 5. Profile for B2B insurance life/pension Sweden 2007

Profile Life/pension B2B segment EPSI 2007 Sweden

Index score

75

70

65

60

55

50

 

It is seen that the image as well as the loyalty (trust) in the company vary widely between insurance companies, while the expectations are more even. None of the insurance providers deliver product and service quality in accordance with what is expected. This gap is especially large for a few of the big banks, generating low satisfaction. Alecta has obtained the highest improvement in their overall scores from the previous year (2006).

Insurance brokers add value to the insurance products offered

In corporate insurance (both general and life/ pension) insurance brokers play an important role in all Nordic countries. Thus, the assessment of the performance by also the brokers, are important in order to understand the functioning of the insurance sector. With a start from year 2007 EPSI has also initiated studies of insurance brokers. The first study was conducted in Sweden. The satisfaction levels are high (74.5 on average). Among the largest insurance brokers Söderberg and Partners and Säkra are on top. Länsförsäkringar are on top also among the customers to insurances through brokerage with If on second place.

The profile for corporate insurances via broker and those directly assigned by the customer are illustrated below.

The added value of brokerage in insurance is here estimated to be between 5 – 10 units on the EPSI scale (highest in the life/pension area).

Graph 6. Profile for B2B insurance through broker and direct relationship, Sweden 2007

Profile by relationship Broker - direct EPSI 2007 Sweden

Index score

85

80

75

70

65

60

 

Product specific rating indices – Vehicle insurance

Within the general insurance final consumer segment we have now extended the annual benchmark study to cover also a special survey of vehicle insurance in Sweden

The Satisfaction indices with the auto vehicle insurance company are given below for the Swedish market. Here also companies focussing on automotive insurance – including the various vehicle brands are included. It is to be noted that these vary quite a bit from the overall scores given for general insurance by the customers. The reason for this is that in the global study of general insurance the respondent is asked to think about the company with which he has the most active relationship. Vehicle insurance does not always come on top in that situation. More often home/ villa-insurance is the first one to refer to by a responding consumer.

From the diagram below we see inter alia that the perceived loyalty differs within wide margins between companies. Volia and Länsförsäkringar have the strongest loyalty for a given satisfaction level (seen by the direction of the curve to the right) while both Folksam and Trygg Hansa show a low perceived loyalty. They have to work hard to maintain the current customer base for any given satisfaction level, as the curve points downwards.

Graph 7. Profile for B2C Vehicle insurance Sweden 2007

Profiles Vehicle insurance B2C EPSI Sweden 2007

80

75

70

65

60

55

Index score

 

Customers are different – Insurance company profiles

Customers are different, exhibiting varying satisfaction levels. Differences in customer profiles may depend on individual (sociodemographic and life-style) characteristics as well as on mis-matches between expectations and deliveries. The aim should certainly be to minimize the number of dis-satisfied customers at the same time as improving the satisfaction for those that are generating the lion-share of profits is rational. Analysis shows that sometimes golden clients have other priorities than average customers. However, all dis-satisfied customers constitute a threat to the recognition of the insurance company. Here also the handling of claims play an important role for satisfaction. All this is

Graph 8. Satisfaction profile by level

possible to assess in the EPSI-approach.

As a standard rule-of-thumb it may be taken that customers with satisfaction below 60 are very dis-satisfied and vulnerable for leaving, and not least due to communicate bad word of mouth about the insurance company. In the same fashion, customers with satisfaction on 75 or above are most loyal and due to spread positive comments about the company. The profile below illustrates the situation. It is taken from the Norwegian general Insurance B2C-study 2007. As is clear from this the aim should be to reduce the percentage of customers with low loyalty for company 1. Their challenge is to try finding some "common features" among these customers in order to devise possible targeted improvement programs.

Spread Customer Satisfaction General Insurance B2C EPSI Norway 2007

Proportion (%)

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0%

0-60 60-75 75-100

Index level

company 1 company 2 company 3 company 4 Other

There is a strong relationship between satis-results for each individual insurance company faction and loyalty as is illustrated by the (B2C-market) in the Nordic region measured following Graph 9, based on an analysis of the separately 2006.

Graph 9. Relationship between Satisfaction and Loyalty in Insurance sector

Customer satisfaction vs Loyalty Propetry insurance B2C 2006 EPSI Rating

Loyalt

90

85

80

75

70

65

60

 

Most customers in the insurance sector do not above. In the following table the percentage of actively consider any alternative to their cur-final consumers considering alternatives to rent company, though the perceived loyalty their current general insurance company are for any given satisfaction level differs widely given. from company to company as was indicated

Table 4. Percentage of final consumers by alternatives seen

Alternative to the current company Denmark Finland Iceland Norway Sweden

No other company at all 22.7 31.1 38.5 32.2 31.2

ONE alternative insurance company 8.1 27.7 22.3 14.3 26.1

2 - 3 alternative companies 19.9 28.4 19.6 27.0 26.5

4 or more alternative insurance companies 25.2 8.1 7.1 12.5 8.1

Do not know 24.1 4.8 12.5 14.0 8.2

Total percentage 100.0 100.0 100.0 100.0 100.0

As is seen from this table, the highest percent-the same time, the percentage of customers age of customers that see one or more compet-answering "I do not know" is highest in Denitors to the current insurance company is found mark. in Denmark, while the lowest is in Iceland. At

Use of the results for Priority effects differ significantly. The illustration is Setting taken from the Danish general insurance B2Csegment. The value "total effect" tells how

From the causal analysis it is possible to trace

much satisfaction will be changed (improved)

the relative importance/effects on satisfaction

if the score on the respective driving latent

and loyalty of changing individual drivers.

variable is increased by one unit.

From the graph below it is illustrated that the

Graph 10. Total effects on satisfaction by driving aspect in model

Total Effect from Drivers to Customer Satisfaction - EPSI Denmark 2007 Driver is increased by one unit

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

 

Total Effect

 

company 1 company 2

company 3

company 4 company 5

company 6

company 7

All others

 

As is seen here, improvements in Product Quality will have the highest effect on Satisfaction – and from that on loyalty, with image as the second most effective. However for some individual companies (like company 6) the most effective strategy would be to focus on image (improve or at least maintain). At the same time it is seen that for company 3, "Service Quality" is a much more effective area for resources focusing than for any of the other considered insurance companies.

The driving aspects indicated above are general in character. For true improvement work it is crucial to being able to be more specific. This is possible by analysing individual manifest questions in the survey. Below, this is illustrated for one of the Finnish companies in the corporate segment. We focus on the latent drivers "Image" and "Value for Money".

Graph 11. Score and relative importance for improvement

Score and Weight Image EPSI Finland 2007

9.0

0.25

8.5

Score Weight

0.23

8.0

0.21

7.5

0.19

7.0

0.17

6.5

0.15

the image of being a the image of the image of offering the image of being a the overall image of reliable and providing excellent good value for professional, state-your insurance transparent overall customer money to the of the-art insurance company?

insurance service? customers in terms company? company? of fees and claim settling?

Score

Weight

It is seen from the diagram that the score on the focus should be on the aspect(s)/variable(s) five different sub-aspects vary rather little that is least expensive to improve. If im(between 7.5 – 8 on the 10 unit scale used in provement is targeted to the two aspects interviews) for this company. The bars refer to "excellent customer service" and "value for the 5 questions in the survey related directly to money perception", the following analysis the Image concept. Neither the relative weight and framework for evaluation may be spec-of either of these vary much. That means that ified.

Effect Calculation – Image

A: Improve the image of providing "excellent customer service" with one unit on 1-10 scale 2007 yields (result 2006 in parenthesis)…

….An increase in overall Image by 2.60 units (2.92 units last year) ….An increase in customer satisfaction by 1.63 units (1.93 units)

B: Improve the image of providing "value for money" with one unit on 1-10 scale 2007 yields …..

….An increase in overall Image by 2.10 units ….An increase in customer satisfaction by 1.30 units

The other important aspect for improving "Product Quality". Here four different quescustomer satisfaction in this situation (for the tions are asked in the survey and the follow-company and segment under consideration) is ing analysis foundation may be specified.

Graph12. Score and relative importance "Product Quality"

Score and Weight Product Quality EPSI Finland 2007

9.0

0.30

8.5

Score Weight

 

 

 

 

0.28

8.0

0.26

7.5

0.24

7.0

0.22

6.5

0.20 the quality of the the quality of the the technical quality of the "overall insurance insurance functions reliability and accuracy the products (functions) quality" of the functions offered (fees, coverage, (standing orders offered (data transfer, that matter for you etc.)? processed in Internet connection, offered by your accordance with overall security system, insurance company?

instructions, accuracy etc.)? of statements, etc.)?

Effect Calculation - ProductQ

A: Improving the quality of the "reliability and accuracy" with one unit on 1-10 scale 2007 yields (result 2006 in parenthesis)…

….An increase in overall Product Quality by 2.94 units (3.14 units last year) ….An increase in customer satisfaction by 1.81 units (1.78 units)

B: Improving the quality of "product functions" with one unit on 1-10 scale 2007 yields ...

….An increase in overall Product Quality by 2.70 units ….An increase in customer satisfaction by 1.66 units

Thus – provided these above aspects can be improved in the eyes of the customers it is possible to obtain a significant increase in satisfaction (and consequently loyalty).

 

Score

With 1 unit each (from 7.9; 7.6; 7.9 and 8.1 respectively) – it is possible others have succeeded with improvements of 1 unit or more during 1 – 2 years time

The customer satisfaction may (theoretically) improve with (up to) 6.4 units. Highly significant indeed!

The above illustration gives some ideas about must be kept in mind that preferences of the how the EPSI analysis model may be used for customers may change over time. The ones prioritizing alternative improvement schemes, expressed here are those obtained in the annual and for successive follow-up and evaluating analyses in 2006 and 2007. effects of undertaken activities. At this step, it

Weight

Conclusions and possibilities for Improving Customer Trust

The Pan-Nordic EPSI studies will be run annually also in the future. The coverage and level of detail will be increased step-by-step. Combinations from the insurance sector with other industries will give insight into the general and specific characteristics of the insurance industry. At the same time the global scores may be used for assessing changes in the market operation and penetration within the sector, country by country. For individual companies and business areas the approach further gives insight for priority setting and evaluation work.

Notes

1 ECSI (1998), Foundation and Structure. In the end of 1990-s the European initiative to stimulate the development of National Customer Satisfaction studies was initiated. It started in the mid 1990s by discussions emerging from experience in Sweden, where a national platform was established in 1989, and from USA, where a national study similar to the Swedish model, commenced in 1994.

2 Based on an explicit account of both sample and model uncertainty.

References

Dermanov, V and Eklöf, J. (2001) Using aggregate Customer Satisfaction Index – Challenges and Problems of Comparison with Special Reference to Russia, Paper presented to the 6th World TQM Congress, Saint Petersburg.

ECSI Technical Committee (1998). European Customer Satisfaction Index: Foundation and Structure for Harmonized National Pilot Projects.

Report, October.

Eklöf, J and Parmler, J. (2007a). The Value of Customer Satisfaction. EPSI-research Working Paper.

Eklöf, J and Parmler, J. (2007b). Calibration coefficients for EPSI benchmark studies – some initial results. EPSI-research Working Paper.

Eklöf, J and Parmler, J. (2007c). Relationship between perceived and actual loyalty based on panel studies in EPSI. EPSI-research Working Paper.

EPSI Rating (2007), EPSI Rating Benchmark report 2006, Göteborg 2007.

Fornell, C., M. D. Johnson, E. W. Anderson, J. Cha and B. E. Bryant (1996). The American Customer Satisfaction Index: Nature, Purpose and Findings. Journal of Marketing 60(4): 7-18.

Ittner, C., D. and Larcker, D. F (1998). Are Non-financial Measures Leading Indicators of Financial Performance? An Analysis of Customer Satisfaction. Journal of Accounting Research. Vol. 36, pp. 1-35.

Kristensen, K., Mørch, L., & Sørensen, H. (2006). Relationship between performance measures and financial results in a large Nordic bank Performance Measurement and Management: Public and Private. (pp. 1001-1017). Cranfield School of Management.

Kristensen, K. and A. H. Westlund (2003). Valid and reliable measurements for sustainable non-financial reporting. Total Quality Management 14(2): 161-170. Kristensen, K., & Westlund, A.

H. (2003, okt.). Performance Measurement and Business Results. Paper presented at the 8th World Congress for Total Quality Management, Dubai

Kristensen, K., Martensen, A., & Grønholdt, L. (2002). Customer Satisfaction and Business Performance Business Performance Measurement - Theory and Practice. (pp. 279-294). Cambridge: Cambridge University Press.

Selivanova, I., A. Hallissey, A. Letsios and J. Eklöf (2002). The EPSI Rating Initiative. European Quality 9(2): 10-25