Are Long-Tenure Customers Less Price Sensitive ? An Empirical Examination in a
John Dawes, Ehrenberg-Bass Institute, University of South Australia
Customer loyalty is a goal for many service organizations. Many marketing initiatives are
implemented to favourably affect customer sentiment and build loyalty. Loyalty can be
manifest as the length of the relationship, which is referred to as tenure, or the breadth of the
relationship, also referred to as share of wallet or level of cross-purchasing. A long-tenure
customer base is said to provide many benefits to the service provider. One of these claimed
benefits is reduced price sensitivity (e.g. Reichheld & Teal 1996). One operationalisation of
this benefit is that the impact of a price increase to a long-tenure customer will be lower than
the impact of a price increase to a short-tenure customer. However, limited empirical
evidence exists as to the accuracy of this belief. This study investigated whether long-tenure
customers are indeed more sensitive or less sensitive to price increases. It used individual-
level consumer records (n=69,672) from a large domestic insurance provider. The data was
sourced from internal company records, and was based on the response to renewal notices
sent to domestic insurance customers. The data included the price change the customer
received, ranging from zero to a 20% price increase from the last policy renewal to the current
one; their length of prior tenure with the provider (one to fifteen years), and whether they
remained as a customer, or lapsed their policy after receiving the renewal notice. Several
other control variables were also included. A series of logistic regression models were used
to test the research proposition. The results give positive support to the proposition that
longer customer tenure is associated with reduced sensitivity to price increases.
Customer Retention, and Pricing Decisions
Customers are increasingly being recognised and managed as assets to the firm (Hogan,
Lemon and Rust 2002). A customer base represents a source of future revenue, both in terms
of repeat-purchases of products currently bought, and potential for future purchases not yet
bought. Indeed it is widely accepted that if there are set-up costs to the firm to attract or
recruit new customers, then it is financially desirable to retain current customers rather than
constantly lose customers and incur the expense of replenishing the customer base. This is
not to underplay the importance of also winning new customers, as some loss in the customer
base is inevitable and this must be replenished just to maintain market share (e.g. Sharp,
Riebe et al. 2002).
Aside from initial attraction and set-up costs, there are other benefits said to accrue from
retaining customers for longer periods of time. Many of these benefits were enunciated by
Reicheld and co-authors (e.g. Reichheld and Sasser 1990; Reichheld 1996; Reichheld and
Teal 1996). The benefits listed by Reicheld include amelioration of acquisition costs;
enhanced overall revenue arising from a longer relationship time period; easier servicing due
to customer learning, more referrals, greater tolerance of higher prices, and less likelihood of
defection in future years. In recognition of the stated benefits of customer retention, many
companies have embraced customer satisfaction and relationship marketing initiatives
designed to keep customers for longer, or increase their share of wallet. Indeed, many of the
possible antecedents of customer loyalty and retention have been examined in the academic
literature, such as service quality (e.g. Zeithaml, Berry and Parasuraman 1996), customer
satisfaction (e.g. Bloemer and Lemmink 1992; Biong 1993; Rust and Zahorik 1993;
Paper presented at ANZMAC conference, University of Otago,
Iacobucci, Grayson and Ostrom 1994; Danaher and Gallagher 1997; Patterson and Spreng
1997; Mittal and Lassar 1998; Ranaweera and Prabhu 2003) – for a recent meta-analysis see
Szymanski (2001); loyalty programs (Sharp and Sharp 1997; Bolton, Kannan and Bramlett
2000), and cross-selling to raise customer switching costs and therefore customer longevity
(e.g. Kamakura, Wedel et al. 2003).
Overall, such work has had a broad focus on initiatives that the firm can implement to
favourably affect customer sentiment, or in some other way, make its customers more
behaviourally loyal. However, it is also the case that the firm may sometimes choose to, or be
forced to, do things that potentially have an unfavourable impact on customer sentiment
and/or behaviour. A prime example of this is a price increase. If input costs to the firm
increase, it must raise prices, otherwise margins will suffer. Naturally, one of the basic tasks
of marketing in this situation is to manage this situation such that the potential impact on the
customer base is minimised. However, it is still the case that price rises represent a potential
threat to the establishment and maintenance of long-term customer relationships and loyalty.
Furthermore, organizations have a powerful profit incentive to ensure their prices adequately
reflect value, and achieve margin objectives. For example, Marn and Rosiello (1992) find
that a one percent gain in price has more positive impact on bottom-line profit than a
commensurate cost reduction. Likewise, in many cases price increases may be difficult to
avoid for either marketers or their customers. Pricing decisions are often decided on the basis
of costs (Zeithaml, Parasuraman and Berry 1985) and so flow-on effects in the form of price
rises will occur if costs rise.
Research Question: are long-tenure customers less sensitive to price increases ?
The question arises, what happens when a service provider raises price ? In relation to the
current customer base, elementary knowledge tells us that price is inversely related to
aggregate demand, so the expectation is that a price rise will cause higher levels of switching,
and/or reduced levels of cross-buying among the customer base. It would use useful to know
which customers are more sensitive to the price increase. One common method of classifying
customers is according to their current loyalty status, for example distinguishing between
short-tenure and long-tenure customers. Do these two customers groups differ in terms of
their sensitivity to price increases ? The answer would be useful from a practical viewpoint as
well as an advance in academic research. For marketers, pre-identification may make
retention programs easier to construct; or assist in designing ‘winback’ programs (Stauss and
Friege 1999; Thomas, Blattberg and Fox 2004). Likewise, this knowledge would be
potentially valuable in brand or corporate valuation. If longer-term customers are less price
sensitive, then two companies with the same number of customers but with differing average
customer tenure arguably differ in monetary value.
Despite its potential importance, there is little evidence on this issue. As stated earlier,
Reichheld (1996) claimed that long-term customers are less price sensitive, but offered no
empirical evidence as to this claim. One notable study of this claim was Reinartz and Kumar
(2000) who investigated whether longer-term customers tended to pay less than newer
customers – and found they did not. A study by Danaher (2002) examined the impact of
different forms of price increases among customers of a telecommunications service, but
tenure was outside the scope of that study. However, a literature search found no studies that
have examined whether price increases as implemented by the service provider have
differential impact on long-tenure versus short-tenure customers.
Would we expect long
-tenure customers to be less price sensitive, or more price sensitive,
than short-tenure customers ? There are opposing arguments for this question. To begin
with, it is a reasonable proposition that long-tenure customers of a provider should be less
sensitive to any price increases implemented by that provider. This may be is partially a form
of ‘selection effect’ - very price sensitive buyers presumably are more likely to migrate
between providers and will have lapsed previously. Secondly, if a customer stays with a
service provider for a long period, this may reflect the fact that the utility provided by the
service provider is high – it offers superiority in some aspect of the product offering which
dampens sensitivity to price. Third, tenure provides more opportunity to cross-sell, which can
build switching costs. Conversely, from the viewpoint of perceived fairness (e.g. Homburg,
Hoyer and Koschate) a long-tenure customer who receives a price increase may feel their
long-term loyalty to the provider has been unrewarded and so they may be motivated to
redress this by seeking alternatives. In addition, Reinartz and Kumar (2000) report evidence
that long-term customers may be more ‘value conscious’ hence less likely to accept higher
prices. Given these opposing rationale, the research question is therefore framed as non-
directional – is the effect of price increases on the odds of lapsing different for long-tenure
compared to short-tenure customers ? I report on an empirical study to address this question.
This study examined whether longer-term customers of a service provider exhibited different
levels of price sensitivity compared to newer customers. Price sensitivity was assessed by
examining the odds of not remaining as a customer (ie ‘lapsing’) arising from a price increase.
Note that lapsing is an imperfect loyalty measure as not all lapsers have ‘switched,’ some
merely fail to renew their annual contract. The effect of price on lapsing was compared
across short-tenure customers and long-tenure customers. It utilised database information
supplied by a large car insurance company. The data included information on the premium
paid by the customer at their previous renewal, as well as the price indicated on an invitation
to renew the premium sent one month before the policy was due to expire. From this
information, the price change to the customer was calculated. Several other variables were
included in the analysis as independent variables: the customer’s age, whether the customer
had multiple policies with the provider, the ratio of the original price paid to the coverage of
the policy, and whether payment was by instalment or an annual payment. Age potentially
confounds the effect of tenure because age and tenure are both time-related variables and are
correlated. Its inclusion therefore clarifies the effect of tenure, while including the possible
effect of age. Customers with multiple policies are likely to have a lower baseline lapsing
rate, therefore this is posited to be related to the dependent variable. The ratio of the price
paid by the customer to the coverage they receive is an indicator of how ‘expensive’ (prior to
the price increase) the policy has been. Customers paying more for a given amount of
coverage could be expected to have higher lapsing rates because they have a heightened
economic incentive to seek alternatives. Finally, use of a ‘payment plan’ could lower the
lapsing rate because this splits the overall price paid into smaller, more affordable payments,
lowering the incentive for the customer to consider alternatives at renewal time.
The research proposition stated in statistical terms is that tenure moderates the form of the
relationship between price and lapsing. Form in this context refers to the slope of the price-
lapsing relationship, which is identified from a logistic regression coefficient. As the
dependent variable is binary (remain / lapse), logistic regression is the appropriate analysis
method. The analysis uses Moderated Regression Analysis (MRA) - (Sharma, Durand and
Gur-Arie 1981), adapted for a logistic model, for identifying moderator variable relationships.
The approach commences with specifying three regression equations to identify if the
coefficients are equal in each case.
2 X 2
2 X 2 +
3X1 * X 2
See Sharma, Durand and Gur-
Arie (1981 p. 295)
Where P is the probability of lapsing. P/(1-P) is the odds of lapsing. In this model X
extent of upward price increase faced by the consumer ranging from 0 to +20 percent more
than their last renewal price. X
is tenure in years, ranging from 1 to 15 years. The other
independent variables are omitted from these illustratory equations for simplicity. If the
coefficient for tenure is statistically significant in (2), but the interaction term in (3) is not,
then tenure is merely a predictor variable. If the coefficient for the price x tenure interaction
in (3) is statistically significant, this indicates that the effect of price on the odds of lapsing is
contingent on the tenure of the customer. A negative coefficient would indicate that tenure
dampens the effect of a price increase, whereas a positive coefficient would indicate longer
tenure heightens the effect of a price increase. The interaction term was constructed using
mean-centred price and tenure variables to reduce co-linearity. The price and tenure variables
used as independent variables were not mean centred. Results are shown below.
Table 1 MRA: Model Comparisons. Dependent variable is odds of lapse v. renew.
Price, + age,
Price, Tenure +
policies & ratio
Price, Tenure, Price x
Tenure, + age, multiple
policies & ratio
Price x Tenure
The interaction term is statistically significant, suggesting tenure moderates the effect of price
The model parameters are all statistically significant at p< 0.05 or less, and the Nagelkerke fit
statistic is reasonable at 0.10. It is difficult to produce a ‘large’ fit statistic from a logistic
model when the variable of interest is far from a 50:50 split – here, the baseline split is 90%
renew and 10% lapse. In this instance, the ‘null’ model can correctly classify 90% of cases
by classifying every case as a non-lapser. Therefore a pseudo R
of 0.10 is quite acceptable.
The coefficient for the interaction term price x tenure in the MRA is statistically significant
8, df 1, p<0.01) even though the log likelihood ratio is reduced by a comparatively small
amount. Given that the coefficient for tenure and the interaction term are both statistically
significant, the appropriate classification for tenure is a
quasi-moderator (Sharma, Durand
and Gur-Arie 1981). That is, it is related to the criterion variable and interacts with the
predictor (independent) variable, namely price. The negative sign of the interaction
coefficient suggests it dampens / weakens the effect of price on the odds of lapsing. That is,
as tenure increases, the impact of higher prices on the odds of lapsing weakens. Based on the
MRA results, I conducted a sub-group analysis to further clarify the role of tenure. The
sample was split into two groups: a short-tenure group and a long-tenure group. Results are
shown below. The mean tenure in years of the short-tenure group was 2.2 years, the long-
tenure group averaged 9 years. Results are in Table 2 below.
Table 2: Sub-group analysis. Dependent variable is odds of lapse v renew.
Group 1: short-
Group 2: long-
The coefficient for price is lower among long-tenure customers
The sub-group analysis further clarifies the moderating role of tenure. The coefficient for
price is markedly lower among long-tenure customers, by approximately half. Likewise the
explanatory power of the logistic model is lower among long-tenure customers suggesting
that price simply has less impact on the odds of lapsing among longer-tenure customers.
We can exponentiate the coefficients (e^
) to gain a sense of how much difference in the odds
of lapsing there are among short-tenure customers compared long-tenure customers when
faced with a price increase. The average odds of lapsing among short-tenure customers is
0.13 (11.5% lapse / 88.5% renew = 0.13). The exponentiated coefficient for price among
short-tenure customers is e^
= 1.036. A 1% price rise therefore raises the odds of lapsing
by a factor of 1.036, which translates to a new lapsing rate of 11.9%. In other words, a 0.4
percentage point increase in lapsing for every 1% rise in price. Among long-tenure customer
the average odds of lapsing is 0.076 (7.1% lapse, 92.9% renew = 0.076). The exponentiated
coefficient for price among short-tenure customers is 1.015. A 1% price rise therefore raises
the odds of lapsing by a factor of 1.015. This translates to a new lapsing rate of 7.2%. In
other words, a 0.1 percentage point increase in lapsing for every 1% price rise. These figures
give some sense of the managerial implications of the results presented here. Larger price
increases would obviously result in quite marked differences in the change in lapsing rates
among short-tenure versus long-tenure customers. These results, albeit based on only one set
of data, suggest tenure does have a statistically and managerially significant relationship with
reduced sensitivity to price increases. Does tenure ‘cause’ reduced price sensitivity ? This is
unknown. It may be that a common factor underlies the relationship. For example, perhaps
the long-tenure customers in this study are inherently less price sensitive. Such causal
questions are difficult to answer. The next step for this research is replication, to test the
generalisability of the tenure – reduced price sensitivity relationship.
Biong, H. (1993) Satisfaction and Loyalty to Suppliers within the Grocery Trade. European
Journal of Marketing 27 (No. 7): 21-38.
Bloemer, J. M. M., and Lemmink, J. G. A. M. (1992) The Importance of Customer
Satisfaction in Explaining Brand and Dealer Loyalty. Journal of Marketing Management 8:
Bolton, R. N., Kannan, P. K., and Bramlett, M. D. (2000) Implications of loyalty program
membership and service experiences for customer retention and value. Journal of the
Academy of Marketing Science 28 (1): 95-108.
Danaher, P. J. (2002) Optimal pricing of new subscription services: analysis of a market
experiment. Marketing Science 21 (2): 119 - 138.
Danaher, P. J., and Gallagher, R. W. (1997) Modelling Customer Satisfaction in Telecom
New Zealand. European Journal of Marketing 31 (No. 2): 122-133.
Hogan, J. E., Lemon, K. N., and Rust, R. T. (2002) Customer equity management: charting
new directions for the future of marketing. Journal of Service Research 5 (1): 4-12.
Homburg, C., Hoyer, W. D., and Koschate, N. (2005) Customers' reactions to price increases:
Do customer satisfaction and perceived motive fairness matter? Journal of the Academy of
Marketing Science 33 (1): 36-49.
Iacobucci, D., Grayson, K. A., and Ostrom, A. L. (1994) The Calculus of Service Quality and
Customer Satisfaction. In Advances in Services Marketing and Management. Eds. T. A.
Swartz, D. E. Bowen and S. W. Brown. Greenwich, Connecticut: JAI Press Inc., 1-67.
Kamakura, W. A., Wedel, M., de Rosa, F., et al. (2003) Cross-Selling Through Database
Marketing: A Mixed Factor Analyzer for Data Augmentation and Predication. International
Journal of Research in Marketing 20: 45-65.
Marn, M. V., and Rosiello, R. L. (1992) Managing Price, Gaining Profit. Harvard Business
Review September//October: 84-94.
Mittal, B., and Lassar, W. M. (1998) Why do customers switch? The dynamics of satisfaction
versus loyalty. The Journal of Services Marketing 12 (3): 177-194.
Patterson, P. G., and Spreng, R. A. (1997) Modelling the relationship between perceived
value, satisfaction and repurchase intentions in business-to-business, services context: An
empirical examination. International Journal of Service Industry Management 8 (5): 414-434.
Ranaweera, C., and Prabhu, J. (2003) On the relative importance of customer satisfaction and
trust as determinants of customer retention and positive word of mouth. Journal of Targeting,
Measurement and Analysis for Marketing 12 (1): 82-90.
Reichheld, F. F. (1996) Learning from Customer Defections. Harvard Business Review 74
Reichheld, F. F., and Sasser, W. E. J. (1990) Zero Defections: Quality Comes to Services.
Harvard Business Review 68 (5, September-October): 105-111.
Reichheld, F. F., and Teal, T. (1996) The Loyalty Effect: The Hidden Force Behind Growth,
Profits, and Lasting Value. Boston, Massachusettes: Harvard Business School Press.
Reinartz, W., and Kumar, V. (2000) On the profitability of long
-life customers in a
noncontractual setting: an empirical investigation and implications for marketing. Journal of
Marketing 64 (4): 17-35.
Rust, R. T., and Zahorik, A. J. (1993) Customer Satisfaction, Customer Retention, and Market
Share. Journal of Retailing 69 (No. 2, Summer): 193-215.
Sharma, S., Durand, R. M., and Gur-Arie, O. (1981) Identification and Analysis of Moderator
Variables. Journal of Marketing Research 18 (August): 291-300.
Sharp, B., Riebe, E., Dawes, J. G., et al. (2002) A Marketing Economy of Scale - Big Brands
Lose Less of their Customer Base than Small Brands. Marketing Bulletin 13: 1-7.
Sharp, B., and Sharp, A. (1997) Loyalty Programs and Their Impact on Repeat-Purchase
Loyalty Patterns. International Journal of Research in Marketing 14 (5): 473-486.
Stauss, B., and Friege, C. (1999) Regaining service customers: costs and benefits of regain
management. Journal of Service Research 1 (4): 347-361.
Szymanski, D. M., and Henard, D. H. (2001) Customer satisfaction: a meta-analysis of the
empirical evidence. Journal of the Academy of Marketing Science 29 (1): 16-35.
Thomas, J. S., Blattberg, R. C., and Fox, E. J. (2004) Recapturing lost customers. Journal of
Marketing Research 41 (1): 31-45.
Zeithaml, V. A., Berry, L. L., and Parasuraman, A. (1996) The Behavioral Consequences of
Service Quality. Journal of Marketing 60 (2, April): 31-46.
Zeithaml, V. A., Parasuraman, A., and Berry, L. L. (1985) Problems and Strategies in
Services Marketing. Journal of Marketing 49 (Spring): 33-46.