why do consumers pay voluntarily? evidence from online music tobias regner and javier a. barria

Why Do Consumers Pay Voluntarily? Evidence from Online Music - T. Regner and J. Barria

Why Do Consumers Pay Voluntarily? Evidence from Online Music

Tobias Regner and Javier A. Barria

[email protected]@imperial.ac.uk

This work was partially supported by the EU-funded projects TIRAMISU (IST-2003-506983) and DANAE (IST-2004-507113). We are grateful to Maija Halonen-Akatwijuka and Osiris Parcero for valuable comments. Finally, John Buckman has been essential for this paper as he provided the

Magnatune data.


Why Do Consumers Pay Voluntarily? Evidence from Online Music

• Analysis of payment behaviour– Alternative online music label: Magnatune– Customers choose price for an album from a given range– How much do they pay, what makes them pay more?

• Do consumers pay voluntarily?– Large data set obtained from Magnatune– Field study instead of data from lab experiments– Quantitative analysis

• Why do consumers pay voluntarily?– Online survey of frequent Magnatune customers– Qualitative/quantitative analysis– What is the underlying motivation for paying more than necessary?


• Niche music label with 200 artists– Slogan: We're a record label. But we're not evil.– Revenues are evenly split between artist and Magnatune

• Various genres, relatively unknown artists• no Digital Rights Management

– Music files are unprotected– File quality and format is up to customer (MP3, WAV, ..)

• Comprehensive pre-purchase access– Free and unlimited streaming as well as online radio service

• Not just 30 second snippets– Free sampling of experience goods facilitates discovery of music

• Variable pricing– Customers are free to pay what they want– Price range from $5 to $18– Recommended price/default value: $8

Magnatune: Music Label and Online Store


Data Set

• Sales numbers of the music label Magnatune www.magnatune.com

• 14,367 observations in total• Data from September 2003 to January 2005• 7,620 different customers• Majority (4,986) purchased just one album• Most albums bought by a customer: 49• On average customers bought 1.86 albums

• No reputation effects possible


• Customer/artist ID• Payment• Date and time of purchase • Dummies

– Gender– Download or CD– Email left or not– Means of payment (credit card, PayPal)– Country– Music genre

• Total amount of purchases of customer • Number of a respective purchase of a customer

Data Set (2)


Descriptive Stats

TABLE 3: DESCRIPTIVE STATISTICS

Variable obs mean median mode st.dev. min max

Payment [$] 14,367 8.197 8 8 2.301 5 24

Total purchases 7,620 1.86 1 1 2.657 1 49

time difference [days]

14.367 14.57 0 0 45.43 0 429

“female” dummy 10,718 0.087 0 0 0.282 0 1

“no e-mail” dummy 14,367 0.044 0 0 0.206 0 1

“CD” dummy 14,367 0.039 0 0 0.195 0 1

“PayPal” dummy 14,367 0.297 0 0 0.457 0 1


Payment over Countries

TABLE 4: PAYMENT OVER COUNTRIES

Country obs. Mean median mode st.dev. min max

Canada 371 8.08 8 8 1.82 5 18

Switzerland 81 9.28 8 8 2.98 5 18

Germany 190 8.21 8 8 2.38 5 22.03

Spain 55 7.67 8 8 1.49 5 12.24

France 163 8.52 8 8 2.09 5 18

Israel 8 6.75 5 5 1.98 5 10

Italy 65 8.67 8 5 3.85 5 22.03

Japan 110 7.74 8 8 2.23 5 16

Mexico 30 5.90 5 5 1.42 5 10

Singapore 13 7.46 8 8 1.51 5 10

UK 366 8.65 8 8 2.36 5 18

USA 3,838 8.18 8 8 2.14 5 18


Summary of Results

• Payment – Only 2,081 purchases (or 14.5%) were at the minimum of $5– Average payment for an album is $8.20– First payment of customers averages $8.36, 2nd payment averages $8.21

• Regression results (significant at (at least) 5%-level)– Anonymous customers tend to pay less– CD buyers tend to pay more– No difference between male and female customers– No country effects after accounting for GDP differences

• Types of consumers– Clustering of the observed transactions based on previous payment

• minimum, default, high (>$12) payment– $5 previous payment, average payment : $5.49– $8 previous payment, average payment: $8.01– > $12 previous payment, average payment: $12.29

– Payments type-dependent, subject-specific α


Payment Patterns of 32 Customers with 15+ Transactions

id6252

0

2

4

6

8

10

12

1 3 5 7 9 11 13 15 17

id6252

id211

0

1

2

3

4

5

6

7

8

9

1 3 5 7 9 11 13 15 17

id211

id960

0

2

4

6

8

10

12

14

16

18

1 5 9 13 17 21 25 29 33 37 41 45 49

id960

id5174

0

2

4

6

8

10

12

1 3 5 7 9 11 13 15 17 19

id5174

Default (with deviations): 20 consumers

(quasi-)

minimum: 4

>AVG: 5

Other patterns (decreasing payment): 3


• Magnatune customers with at least 10 transactions• Anonymous participation• 236 responses out of 520 invitations sent out• Seven open-ended questions

– Benefits of purchasing music at Magnatune– Own payment (less, around above recommended price) and reasons for it– Estimate of customers’ average payments at Magnatune and reasons

• Eleven multiple choice background questions– Gender– Age– Country– Personal income– Educational background– Actively making music– Donations to charity– Own consideration of others’ decisions/opinions– "It is important to me what others think of me." – Reciprocity (positive/negative)

Online Survey


• Clustering of textual responses into categories– Magnatune specific - technical 45.26%– Support of the artist 44.83%– Magnatune specific - idea/business 41.81%– Magnatune specific - music 30.17%– No DRM 29.74%– “Try before buy" 21.98%– Online music general 9.05%– Guilt 2.16%

• Probit regressions– Dependant variable: Own payment less/around/above recommended price– Explanatory variables

• Category entries• Survey scores (reciprocity, “own consideration”, “importance of others

opinion”, ..)• Gender, country, income, age, education, “actively making music”, “donations”

Qualitative/quantitative analysis


• Own payment is less/around/above recommended price …– Less 13.6% (32)– Around 59,7% (141)– Above 26.7% (63)

• Regression results (***=1%, **=5%, *=10%)– Own payment is above recommended price explained by

• Score of positive reciprocity **• “Try before buy“ **• Middle income regions ***• R²=0.26

– Own payment is around recommended price explained by• Own consideration of others’ decisions/opinions *• R²=0.14

– Own payment is below recommended price explained by• Low income regions ***• Magnatune specific – technical **• R²=0.32

Key Preliminary Results


• Consumers pay voluntarily at Magnatune– Average payment $8.20– Only 14.5% of purchases were at the minimum of $5– Payments appear to be type-dependant

• Why do they pay voluntarily?– Preliminary evidence that positive reciprocity drives behaviour

• Kind behaviour by Magnatune (free streaming etc.)• Score of positive reciprocity significant at 5%-level• “Try before buy“ category significant at 5%-level

– Own consideration of others’ decisions/opinions• Score affects “around”-payments at 10%-level

– Income– No significant effects for gender, education, “actively making

music”, “donations”, “guilt”, “fairness”

Conclusions

why do consumers pay voluntarily? evidence from online music tobias regner and javier a. barria

Documents

online storeniche music

online musictobias regner

music label magnatune

online radio servicenot

record label

magnatune data

respective purchase

email dummy14