Maximising ARPU

Price Elasticity

Coleago Consulting Ltd

Martin Duckworth, Director
14 March 2005
Contents

W hy is elasticity important?

W hat is elasticity?

How can we measure elasticity?

Summary

2
Why is elasticity important

3
Pricing is clearly one of the key determinants of ARPU

Pricing feeds through directly into ARPU

Creative pricing can stimulate demand

Pricing is one of the key tools in any marketing strategy for

customer acquisition and retention

4
Pricing is driven be a range of factors

Competitor

Behaviour

Customer
Pricing Costs
Behaviour

Regulation

5
Each time we change prices we need a to go through a check list

Is the new price profitable

 Is price above incremental cost?
 Is it above incremental cost + mark up

W hat will be the competitive response?

W hat is the revenue (ARPU) impact of a price change?

 What volume change can I expect?

Is the price consistent with regulation?

6
We need a set of tools to address these questions

Profitability Cost Model

Competitor analysis/
Competition
Game Theory

Elasticity
Revenue Impact
estimates

Regulation Compliance

7
What is elasticity?

8
Elasticity definition

Demand price elasticity is the responsiveness to demand to a

change in prices:

percentage change in demand

Ed  
percentage change in price

9
Elasticity is a model of customer behaviour

Customers behaviour is the result of a large number of

purchasing decisions

Customers purchase when the value of the product to the

consumer, for example a call, exceeds the price

If the price changes, consumption will also change

 A reduction in price will result in more calls with value > cost

Elasticity is an expression of the reaction of a customers to an

incremental change in price

10
Consumer behaviour in mobile telephony is complicated

Customers understand the value of calls

 A repeat purchase

However customers awareness of the price of calls is limited

 Customers have limited awareness of prices
 Tariff plans hide the price of an individual call
 Feedback on the cost is delayed

For innovative services both the pricing and the value of the
service may be unclear

11
For many consumers the overall bill is more important than
individual prices

Customer may see value on a individual decision basis...

 “Is this call worth more to me than its cost?”
“Value”

Service
Elasticity

... or on an aggregate basis

 “Is my current usage worth more to me than my monthly bill?”
“Budget”

Aggregate
Elasticity

12
Setting prices according to elasticity can increase revenues without
an overall increase in prices

Increase prices for the most inelastic customers and services

 Little impact on volume so that the price increase flows through to
revenues

Reduce prices for the most elastic customers and services

 The effect of price reductions is offset by increases in volume

“Ramsey pricing” states what the theoretically optimal pricing

is

13
Elasticity may be in conflict with other objectives

The most inelastic customers tend to be the most valuable (e.g.

business users)
 There is often a desire to reduce prices for customer acquisition and
retention

Inelastic customers may be less valuable (e.g. low spending

residential users)
 There may be a desire to raise prices to increase profitability

Prices changes may provoke a competitive response

 Price reductions can spark a price war
 Price increases can lead to increased churn if not matched by competitors

14
Aside : Elasticity in regulation

Regulators are starting to question price discrimination

 Different prices for services which appear to have similar costs

A number of services are being investigated

 Mobile call termination
 On net calls
 Roaming calls

Elasticity provides an argument for price discrimination

 Theoretically efficient “Ramsey” prices
15
Measuring Elasticity

16
Measuring elasticity approaches

Traditional “macro” approach

 Model the relationship between demand and prices for the market
as a whole over a period of time

A “micro” approach
 Response of individual customer segments to a given price change

Other approaches
 Customer surveys
 “Cross sectional” approach to analyse differences in demand
between different groups of customers who have different prices

17
The first step is finding a consistent time series for volumes and
prices...

UK Mobile Retail Minutes and Revenues

70,000 0.45
Minutes (millions)

0.4

Revenue/minute
60,000
0.35
50,000 0.3
40,000 0.25 Mmitts
30,000 0.2 £/minute
20,000 0.15
0.1
10,000 0.05
0 0
19 /97
19 /98
19 /99
20 /00
20 /01
20 /02
20 /03

4
/0
96
97
98
99
00
01
02
03
19

Year Source : OFCOM (excludes ‘3’)

18
Volume changes do appear to mirror price changes...

UM Mobile Retail Minutes and Revenues

80%
Annual Change

60%
40%
Volume change
20%
Price Change
0%
-20%
-40%
8

4
/9

/9

/0

/0

/0

/0

/0
97

98

99

00

01

02

03
19

19

19

20

20

20

20

Year to:

19
... And there appears to be a clear correlation between changes in
price and change in demand

UK Mobile Retail Minutes and Revenue

100%

80%
Volume change

Volume change
60%

40% Expon. (Volume

change)
20%

0%
-30% -20% -10% 0% 10%
Price change

20
However the traditional “macro” approach is of limited
value

Correlation does not imply causality

 Price may not be the key determinant of changes in demand
 Price changes may be the result of an increase in demand and
hence reduction in costs

As neither pricing nor customers are homogeneous, results

cannot necessarily be applied to an individual segment
 There is a significant amount of price discrimination
 Elasticity is likely to vary between customer segments

Historic relationships may no longer be relevant

21
Similarly for fixed to mobile calls volume changes do appear to
mirror price changes...

UK Fixed to Mobile Retail Minutes and

Revenues

80%
Annual Change

60%
40%
Volume change
20%
0% Price Change
-20%
-40%
8

4
/9

/9

/0

/0

/0

/0

/0
97

98

99

00

01

02

03
19

19

19

20

20

20

20

Year to:

22
... Although the correlation is far less clear

UK Fixed to Mobile Retail Minutes and

Revenue

100%
Volume change

80%
Volume change
60%
40% Expon. (Volume
change)
20%
0%
-40% -30% -20% -10% 0% 10%
Price change

23
For SMS there is little evidence of elasticity

UK Mobile SMS and Revenues

2000 0.1
SMS (millions)

Revenue/SMS
1500 0.08
0.06 Messages
1000
0.04 £/SMS
500 0.02
0 0
0

4
/0

/0

/0

/0

/0
99

00

01

02

03
19

20

20

20
20

Year Source : OFCOM (excludes ‘3’)

24
Micro approach

Identify relevant price changes

 May be temporary promotions or permanent changes

Extract information on usage for data warehouse for a period

covering the price change

Time series analysis to remove the impact of seasonality, any

underlying trend and to filter out noise

25
The results of price changes may be masked by other sources of
volatility

Price
Demand

Time ---->

26
Time Series Analysis must be used to “decompose” demand in order
to extract the elasticity effect

Trend
Seasonality
Noise
Elasticity
Demand

Time --->

27
“micro” approach

The micro approach provides targeted and timely elasticity

estimates
 We can use segmentation consistent with the marketing strategy

W e need to identify appropriate price changes

Also need a long enough time series to enable us to extract

the effect of the price change from other sources of volatility

28
Summary and Conclusions

29
Summary : How do we use elasticity?

Efficiently price to maximise ARPU and profitability

 Reduce prices for elastic services to stimulate demand
 Increase prices for the most “inelastic” services to increase revenues

Effectively compete
 Understand the revenue impact of competitive price cutting

Financial forecasting

Justify prices to regulator

 Argue that price discrimination is “efficient”

30
Summary : How can we measure elasticity?

Traditional approach based on the whole market may provide a

first estimate
 Results may be too general for targeted pricing actions

The increasing availability of detailed data allows us to carry

out more relevant analysis
 For individual segments
 For individual price changes

31
Conclusions

Elasticity estimates are an important part of the marketing

toolkit

Adjusting prices to reflect elasticity can boost ARPUs

W hile estimating elasticity is difficult, it is worthwhile

32
www.coleago.com
Martin Duckworth, Director
martin.duckworth@coleago.com
+34 679 472760