You are on page 1of 13

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/263541979

The Competitive Landscape for Leisure: Why Wide Appeal Matters

Article in International Journal of Market Research · January 2014


DOI: 10.2501/IJMR-2015-019

CITATIONS READS

17 343

4 authors:

John Scriven Diana Perez-Bustamante Yabar


University of South Australia King Juan Carlos University
24 PUBLICATIONS 312 CITATIONS 20 PUBLICATIONS 297 CITATIONS

SEE PROFILE SEE PROFILE

Maria Clemente Dag Bennett


London South Bank University London South Bank University
5 PUBLICATIONS 136 CITATIONS 39 PUBLICATIONS 550 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Dag Bennett on 28 February 2015.

The user has requested enhancement of the downloaded file.


International Journal of Market Research Vol. 57 Issue 2

The competitive landscape for leisure


Why wide appeal matters

John Scriven
Ehrenberg-Bass Institute
Diana Perez-Bustamante Yabar
Rey Juan Carlos University
Maria Clemente
London South Bank University
Dag Bennett
London South Bank University

This article reports the results of an analysis of participation across a range of


leisure activities in the UK. This work follows that of Chris Hand and Jay Singh
in the January 2014 issue of IJMR, which analysed participation and partitioning
in the UK betting market using the same UK government DCMS Taking Part
database. Our paper uses a duplication technique, widely used in other consumer
goods markets, which gives a clear understanding of the polygamous portfolio
structure of leisure choices, revealing for the first time how choice processes for
free time activities are similar to those for other consumer goods. The results
show that leisure activities of all kinds compete for the free time choices of all
consumers. The market has some partitioning – for example, ‘cultural’ activities
attract more overlapping customers than expected. However, those in this group
do not participate in these activities to the exclusion of more populist ones, they
are at least as likely to participate in those too. We contrast this with the extant
literature, much of which suggests more marked segmentation between leisure
activities. This has major implications for the purveyors of competitive offerings;
in particular it means that whether marketing ‘cultural’ or ‘leisure’ activities,
strategies that emphasise reaching the largest possible number of occasional
customers are most likely to succeed.

Introduction
Leisure is a major activity: according to the OECD (2009), in developed
countries people spend 20–25% of their time on leisure activities,
significantly more than they spend in paid employment. And this free

Received (in revised form): ?? ?????? ????

© 2015 The Market Research Society 1


DOI: 10.2501/IJMR-2015-000
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

time is growing, both in the short and long term (OECD 2009; McLean to participate in activities outside the group. For example, are opera fans
& Hurd 2012). Working hours per year in the industrial West reduced more likely to also go to the ballet and be less likely to watch television?
from around 3,500 to about 2,000 from 1840 to the present (McLean & How much do consumers choose different activities? Do consumers have
Hurd 2012). This presents consumers with growing possibilities for leisure wide-ranging portfolios or do they concentrate on a few specific types of
activities, and presents new opportunities for the industries that cater to activity? Do consumer types or activities segment leisure time, each activity
them. competing more strongly with limited alternatives, or do all activities share
The traditional framework for analysis of leisure activities has mostly usage proportionately across the alternatives? Could activities operate as
been sociological in nature. The giant in the field was Pierre Bourdieu, a single market that competes for consumers’ free time in much the same
who pioneered concepts such as cultural and social capital to reveal the way as brands compete in goods and services (Ehrenberg et al. 2004)?
motivations and dynamics of power relations in social life (Bourdieu The results help us to understand how activities compete for free
1984). His research showed that possession of cultural capital is closely time and what this means for marketing of various activities. Across the
predicted by social origins, although education (and wealth) offers the huge range of leisure activities competitive purveyors take a variety of
opportunity to acquire social status. Such concepts continue to shape approaches to attract customers. In the case of culture and arts activities,
sociological views of consumer behaviour and also managerial decision marketing strategies generally aim to promote appreciation of the arts
making, especially in ‘cultural’ fields such as museums and the classical through education (Colbert 1997) thereby fostering the acquisition and
performing arts. While this approach is useful in explaining why people growth of cultural capital. However, recent trends suggest that the more
choose particular leisure activities (preferences are derived from social successful strategies have shifted emphasis from high culture to more
position and desires for mobility, and so on) it is less useful in explaining popular cultural appeal (Mencarelli et al. 2010). For example, many of the
what they actually do. Later work, notably by Richard Peterson, takes a more successful museums are those that create more interactive, involving
more bottom-up approach based on empirical evidence of broad-based experiences, employ new technologies and add an entertainment factor to
behaviour, while maintaining a sociological perspective. education. In an increasingly competitive market, museum managers who
Our approach uses a scientific framework that seeks to understand identify and analyse what drives changes in visitors’ choices, and have a
behaviour in the context of theories that have been subjected to rigorous better understanding of customer motivations and their relationship to
testing. While theories provide interpretation, they must be capable of cultural products and services, may be able to raise the appeal of visiting
and subjected to repeated empirical verification. Such testing may at some these types of institution (Mencarelli et al. 2010). Changes in museum
point prove the theory false, or establish a boundary condition that limits marketing also reflect changes in the director role (Gilmore & Rentschler
the applicability of the theory. With regard to leisure activities, we make 2002), which increasingly call for directors to add the orientation and skill
the a priori assumption that choice behaviour for these types of activity sets of marketing managers to their custodial skills.
will be like that for other consumer choices – the reverse of assumptions This popularisation of high culture is somewhat at odds with the
made in the sociological tradition. Thus, for leisure activities, we are testing traditional view that there is a discerning elite with well-refined tastes
whether the behavioural data stand up to the well-established theories (Glenn 1969; Bourdieu 1984) who attend ballet or theatre and eschew
of consumer behaviour derived from a vast number of examinations of popular pursuits such as watching TV or going to the pub. While this
frequently purchased consumer goods categories. This is a pioneering study view has anecdotal history and some academic pedigree (Bourdieu 1984),
that advances our understanding of leisure choices beyond its sociological more recent work by Peterson (1992, 2005) shows that this view no
roots and places it in a broader context of consumer choice making. longer fits with behavioural data. He therefore proposed an alternative
This paper examines the user base of various different leisure activities hierarchy using the terms omnivore and univore for those who participate
from the DCMS Taking Part study in the UK (described later), in in many activities and those who participate in few. This begs the question
particular to discover whether certain activities attract unique participants of whether there is segmentation or partitioning in today’s market for
and whether activities group together in partitions whose users are much leisure products. Is there evidence in the patterns of leisure usage of
more likely to participate in other activities within the group and less likely markedly different types of audience for certain activities that might

2 3
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

indicate potentially aspirational paths to others? Or are the choices of & Wright 2013) performing arts (Hand 2011), restaurants (Lynn 2013)
individuals so diversely motivated that there is little one can conclude and betting (Hand & Singh 2014).
about potential for participating in one activity through knowledge of One feature of a duplication analysis is that it is straightforward to
participation in another? Our findings across a wide range of popular and perform, often with data that are already available (as this dataset was
less popular activities build on previous work to show that the market for to the authors) and requires no complex statistical procedure. It is also
leisure activities has a competitive structure dominated by the popularity easy to see the structure of relationships in the data. If the analysis reveals
(penetration) of the different leisure offerings. This is a fundamentally partitions, then there are almost certainly some market segmentations
new way of looking at leisure choice behaviour that leads to profound based on user or usage characteristics. The nature of the segmentation
managerial implications for participants in the marketplace. may be obvious from the characteristics of the items in the various
partitions. Further segmentation analyses may then be unnecessary,
although it is often helpful to carry out a simple demographic analysis
Objectives
to check if a usage-based partition corresponds with a particular user
This study examines the user base of leisure activities in the UK, with profile. Any partitioning revealed allows one to proceed to segmentation
the aim to discover whether participants in particular activities tend to analysis with a strong hypothesis of the nature of the user segmentation.
be more or less likely to engage in other forms of leisure activity. We Conversely, if there are no partitions, then there are almost certainly no
specifically address whether consumers of cultural activities avoid more segmentations worth worrying about. In which case it is probably wasteful
popular activities rather than adopt omnivorous patterns of activity. to spend time looking for them (Kennedy & Ehrenberg 2001). In short, a
Our expectation is that there will be a great deal of omnivorousness or duplication analysis is a quick and simple method to reveal the big picture
polygamy across all consumers, and that any partitioning that exists will of competitive structure in any market.
be marginal not exclusive, following the patterns of choice established in
many consumer goods.
The data source
This analysis considers culture and leisure activities as a market in
which different activities compete for consumers’ free time in much the Taking Part is a continuous national survey of leisure, culture and sport
same way that brands compete in other categories (Ehrenberg et al. among those living in a representative cross-section of private households
2004). The Duplication of Purchase Law, described later, establishes the in England. The Department for Culture, Media and Sport (DCMS), in
level of sharing between items of choice (in this case, free-time activities) partnership with its non-departmental public bodies, commissioned the
for un-partitioned markets, i.e. if there are no groupings of activities that survey. We thank them for making data freely available through the UK
particularly share users with one another and share less with users of other data archive (DCMS 2010). Launched in mid-July 2005, during the first
activities. Using these benchmarks, it is then a simple exercise to confirm year 28,117 interviews were conducted by BMRB with adults (16 or over).
whether or not the observed level of sharing is in line with the expected, This paper uses the 2009/10 data comprising 6,097 adult interviews.
theoretical level or whether it is more or less than this normal level, The first-year data were also analysed, with similar findings that are
indicating partitions or segments in the market. The duplication process not reported here. The stability of the results in two separate surveys
was originally developed to explore competition in magazines (Agostini strengthens the reliability of the findings. The DCMS issues annual and
1961), then extensively in television (e.g. Goodhardt 1966; Ehrenberg other reports (DCMS 2011), which we draw on here. The aforementioned
& Goodhardt 1969; Goodhardt et al. 1987) and further to all kinds bodies bear no responsibility for the analysis and interpretation in this
of packaged consumer goods (Uncles et al. 1995) from ‘soap to soup’ paper.
(Ehrenberg et al. 2004), retailers, developed and developing markets, The aim of the survey is to improve the current knowledge base of
automobiles, gasoline and others (see Ehrenberg et al. 2004 for extensive users and non-users by gathering quality-assured data on participation,
references). Recent published examples include beer (Dawes 2008), attendance, attitudes and related factors across the many sectors covered
automobiles in Thailand (Bennett & Graham 2010), radio channels (Lees by the DCMS – for example, arts, sports, museums, broadcasting,

4 5
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

gambling and volunteering. It is a rich data source, of which we use are much more popular than others – television watching, for example,
just a small part: the responses for the activities prompted in the study attracts nearly 90% of all respondents, while playing a musical instrument
questionnaire as ‘things you do in any free time you have’. These are gets 10% participation. Perhaps not surprisingly, pursuits that require
the activities itemised in the following tables. The distinction between less effort are the most popular, such as watching television, spending
free time and leisure is a complex discussion, going back to Aristotle, time with friends and family, listening to music and shopping. Activities
which it is inappropriate to review here. In brief, free time and leisure requiring more effort tend to be less popular. It is in these categories that
are not synonymous – for example, Stebbins (2007) defines leisure as serious leisure, which Stebbins (2007) characterises as having a level of
un-coerced, satisfying activity engaged in voluntarily, nested within free involvement akin to the commitment required for career work, is more
time, time free of unpleasant obligation. The distinction is not crucial to likely to be found, and as Stebbins (2007, p. 38) remarks:
the analysis here.
… it is evident that casual leisure is anything but marginal. Thus, far more people
participate in it than in serious leisure and many of the interviewees in my studies
Finding the structure in sharing ‘free time’ activities [of serious leisure participants] pointed out that they also enjoy and therefore
value their casual leisure.
People allocate their free time across many activities – on average nearly a
dozen separate activities (Table 1). On its own, this finding suggests that When we look at the number of activities participated in by participants
consumers exhibit surprisingly broad choice behaviour. Some activities of each activity, we see almost the opposite ranking from penetration: the
participants in less popular activity do more activities (the correlation
is –0.89). This pattern of the users of the most popular items of choice
Table 1 Participation and portfolio size (UK leisure activities 2009/10) sharing less with other items (i.e. having smaller portfolios) is often seen
in other contexts (e.g. brand choice), and is a phenomenon known as
Activity Penetration (%) Average no. of activities
‘Natural Monopoly’ (Ehrenberg et al. 2004). The pattern occurs generally
Watch TV 88 10.1
Spend time with friends/family 85 10.5
because the leading item of choice is more likely to attract light users of a
Listen to music 76 10.8 category and heavier users duplicate their activities more. Thus in this case
Shopping 71 10.8 those with little free time (light users) are perhaps more likely to engage in
Read 68 11.0 the most popular casual activity (watching TV, spending time with family)
Eat out at restaurants 68 11.4 and less likely to engage in more serious pursuits. Nevertheless, there is
Days out or visits to places 64 11.6 much omnivorousness among the whole population.
Internet/emailing 54 11.6 The next question is what other activities do the participants of each
Gardening 53 11.2 activity take part in? Are users of one popular leisure activity more likely
Sport/exercise 52 11.6 to use other popular leisure activities, and users of less popular cultural
Go to pubs/bars/clubs 45 11.7
activities less likely to use popular activities (and therefore more likely
Go to cinema 44 12.2
to use other cultural activities, since we have shown they have a larger
Theatre/music concerts 43 12.5
DIY 38 11.9
portfolio)? We examine these basic premises of segmentation theory using
Visits to historic sites 36 12.9 a duplication analysis.
Visit museums/galleries 33 13.0 A duplication analysis is a contingency table that cross-tabulates
Play computer games 22 11.9 participants in one activity with those who also participate in each other
Arts and crafts 22 12.7 activity. It is presented with the rows and columns transposed compared
Play a musical instrument 10 12.7 with the more usual way of presenting a cross-tab, and with horizontal
Average 51 11.7 percentages. The subgroups of users of an activity are in the rows and the
Source: authors based on Taking Part Data other activities they use are in the columns, as in Table 2.

6 7
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

Musical instrument Both rows and columns in Table 2 are arranged in descending order of


10

13
11

10
12

12
10

1
12
12
12
13
10
13
12
13
16
13
15
16
17
17
the incidence of the activity in the population in total (i.e. descending order
of penetration of activity, from watching TV by 88% of the population
Arts and crafts

37
22

28
24

22
25

25
24

3
27
25
28
27
28
26
23
27
31
29
35
38
27

Computer games to playing a musical instrument by 10%, as in Table 1). This order, and

37
23

26
23

22
25

26
24

1
23
25
25
33
21
26
29
30
24
27
25
25

27
the transposition of rows and columns, are two of the data reduction
Museums/galleries steps (Ehrenberg 1982) that greatly help in being able to see the patterns
34

45
37

33
39

38
37

6
42
42
45
43
41
44
40
47
54
43
67

37
58
52
and hence the structure in the data without resorting to complex maths.
Similarly the penetrations are shown in the row next to the average
Historic sites 37

48
40

36
41

42
40

6
44
46
50
45
47
47
43
48
57
50

73
40
58
52
DIY duplications so that one can easily see the very high correlation.
40

46
41

38
41

45
41

1
41
44
46
47
54
47
46
44
46

53
50
47
51
49
Glancing along the rows, we can quickly see that people who take
Theatre/music concerts part in one free-time activity also take part in all the other activities, in
44

55
48

43
50

50
48

5
52
56
55
53
51
55
55
63

51
68
70
47
62
68
Go to cinema
varying proportions. For example, reading along the first row of Table 2,
of those who watch TV in their free time, 87% of them spend time with
45

54
47

44
50

51
49

3
49
55
53
58
45
55
58

64
50
58
62
58
54
58
Pubs/bars/clubs family/friends, 77% listen to music, 74% go shopping, and so on to the
48

53
49

45
51

53
50

1
48
56
53
56
48
57

60
58
55
54
54
59
48
56
last column, which shows that 10% of them play a musical instrument.
In row 2, of those who spend time with family/friends, 91% watch TV
Sport/exercise
53

61
55

52
56

60
54

1
57
60
61
65
57

65
66
66
63
68
68
62
62
67
Gardening (first column) and 11% play a musical instrument (last column), and so
55

59
55

53
55

61
57

–2
59
59
61
54

58
56
54
62
74
69
64
50
68
53
on …
Internet/emailing However, if we look up and down the columns of Table 2, we can see
55

65
58

54
60

63
59

2
59
63
63

55
68
67
72
67
66
68
69
80
66
72
Duplication of participation in UK arts and leisure market 2009/10

that the numbers in any column are rather similar to one another. Patterns
such as this tend to be more obvious when the similar numbers occur in
Days out
66

76
70

64
71

74
70

2
71
77

74
74
76
75
78
82
77
89
88
71
83
75
Eat out columns rather than rows, because the eye scans down a column more
70

79
73

68
74

78
75

0
75

81
79
75
78
83
85
88
78
86
86
75
79
79

easily than it does across the row (Ehrenberg 1982). That is why the table
Read is presented this way round, to emphasise the similarity of the numbers in
70

76
72

68
74

79
73

–3

75
76
75
76
75
73
76
82
73
84
86
71
85
78

Shopping
any one column.
74

77
75

71
76

82

–5
76
79
78
78
77
74
78
81
79
76
78
79
77
80
71

Interpreting the patterns


Listen to music
77

84
80

76

88
81

–4
82
82
83
84
79
82
85
86
88
81
87
89
86
87
91

Friends/family
In Table 2, it is obvious that the numbers in any column are quite similar
87

91

85
89

98
89

–7
89
92
93
91
89
90
91
92
93
90
94
94
89
93
89

to one another, while the numbers differ markedly from column to


Sources: authors based on Taking Part Data

91
91

88
90

103
93

Average – predicted duplication –12


91
91
91
90
92
90
93
92
91
92
92
91
93
90
86

Watch TV
column – this is made clear from observing the averages. Furthermore,
the numbers in the columns tend to decrease systematically as we move
% of participants in each FREE

Spend time with friends/family


also do each other activity (in

from left to right – again, this is evident in the decreasing column average
TIME ACTIVITY (in rows) who

Days out or visits to places

from left to right. Whichever group of people we consider that have one
Play a musical instrument
Visit museums/galleries
Theatre/music concerts

Predicted duplication
AVERAGE duplication

free-time activity (any of the rows), about 90% watch TV (in the first
Go to pubs/bars/clubs

Play computer games


Eat out at restaurants

Visits to historic sites

column), while about 13% play musical instruments. There are some
Internet/emailing
Listen to music

exceptions, which we shall come to, but these are small in the context of
Arts and crafts
Sport/exercise

Go to cinema

Penetration

the whole table.


Gardening
Shopping
columns)
Watch TV
Table 2

Now, because of the structure we have imposed, we can interpret the


Read

DIY

main pattern easily as follows:

8 9
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

• the proportion of each free-time activity group who spend time doing be estimated in this case as there is no consistent frequency variable
another specific activity is roughly constant available in the data.
• this proportion decreases systematically with the size (penetration) of In summary, each free-time activity shares users more (and therefore
that second activity. competes more for attention) with popular activities, and less with
minority activities.
The duplication law and D coefficient
Deviations from the pattern
The average duplication for each column in Table 2 (the % ‘who also spend
time’ on a particular free-time activity) compared with the penetration of Of course, there are other individual deviations from the main pattern in
each free-time activity (the % who spend time in the population) shows Table 2. For instance, where the row of ‘play computer games’ intersects
a very high correlation. Where duplication is high, penetration is also the column of those ‘who also do gardening’, we see that only 50% of
high (in the left-hand columns), and where one is low, the other is low those who play computer games also garden, compared with the column
(the right-hand columns). In fact, the correlation is almost perfect, with average of 59% of participants in each activity who also garden. And
a coefficient of 0.99. For each activity, the average duplication is always in the ‘visit museums/galleries’ row, we see that 70% of those visiting
about 1.1 to 1.2 times its penetration. museums/galleries go to theatres/music concerts, much higher than the
Broadly, this can be extended to the relationship between individual column average of 55% of each activity that also go to theatres/music.
activities. We can summarise the relationship by saying that the percentage What the analysis technique has done is to make such variations clear –
of people that in their free-time activities take part in some activity X they stand out from the main structure. Now that the duplications analysis
who also take part in specific activity Y is approximately equal to some has revealed them, the task then becomes one of determining whether
constant (1.16 in our example) times the proportion of the population they are sufficiently large to warrant specific comment, if they show any
who spend time on specific activity Y (specific activity Y’s penetration). sub-pattern within them and, if so, what this might mean. However,
It turns out that this relationship is found consistently in market after without first knowing about the main patterns and organising the data
market; so consistently that it is known as the Duplication of Purchase so that they can be clearly seen, these important questions are themselves
Law (Ehrenberg 1988). obscured.
The D value of 1.16 is the average of the duplication percentages (i.e. In Table 3, to emphasise the deviations, we have replaced each cell
average of the column averages) divided by the average of the penetrations. entry with the deviation of the raw value from the column average, and
The expected value (bY|X) of the duplication of those who participate in highlighted all those that deviate by ±5 or more points from the column
activity X who also participate in activity Y is thus: average (bold for positive and bold italics for negative deviations).
Choosing ±5 points is to a degree arbitrary, in the same way as choosing
bY|X = D * bY a 95% or 90% confidence interval, and is chosen to highlight deviations
that are both significant and meaningful.
(Scriven & Danenberg 2010). These predicted duplications are shown as The positive deviation (Observed > Average) indicates that an activity
the penultimate line of Table 2. would be selected more than expected by users of another activity
A consistent small sub-pattern is that duplications are a bit lower than and, if the deviations were negative, the activity in question would be
predicted when penetration is very high: the deviations in the last row chosen less than expected (O < Average). We could just highlight these
of the table are all negative for the first five activities (i.e. most popular) deviations in the duplications in Table 2, but Table 3 shows most clearly
and positive for the rest. This is a general failure of the simplified the patterns of over- and under-duplication between the activities, and
Duplication Law, approximating a relationship that is not quite constant how the activities group together to indicate partitioning in the overall
(Ehrenberg 1988, p. 178). A more complex model, like the Dirichlet leisure market.
(Ehrenberg 1988), would account for this deviation with size, but cannot

10 11
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

Arts and crafts


Seeing partitions clearly

–6
–4
–3
–4
–1
–2
0
–2
–1
–5
0
1
–1
–1
9
3
7
10

28
galleries The order of free-time activities in Table 3 is changed compared to Table
2, grouping the activities in what we have labelled everyday, venue,

–10
–8
–6
–8
–3
–3
1
–1
3
–5
–4
–1
–2
–8
8
10
23

13
45
Cultural

Museums/

Historic sites home, solo and cultural groups as indicated by the grey shading. These

–10
–8
–6
–8
–3
–2
3
–1
0
–5
–1
2
–3
–8
5
9

25
10
48
concerts
groupings, or partitions, emerge from the data as a consequence of the
–11 way that various activities share participants with one another. The labels,
–8
–5
–8
–3
0
0
–1
8
0
–5
–4
–2
–8
13

12
15
7
55
Theatre/
instrument which we discuss next, are ours based on the activities in the groups.
Partitions occur where groups of activities duplicate more than expected
–3
–2
–1
–3
–1
–1
–1
0
0
–1
–3
0
1
4

3
2
3
4
13
Musical

with one another and/or less than expected with other activities. The table
games
Solo

–3
–3
–1
–2
–3
–2
–2
0
3
3
–5
1
7

11
–2
–2
–2
1
26
Computer
emailing is arranged so that activities that show excess duplication between each of
–10
–7
–5
–6
–6
–2
–2
3
7
2
–9
1

15
8
2
3
4
1
65
Internet/
them are grouped together. These groupings are now discussed, starting in
DIY the bottom right-hand corner of Table 3 with the cultural activities group
–6
–5
–5
–5
–5
–2
0
1
–2
0
8

1
1
3
0
7
4
5
46
Home

and moving up the diagonal to the top left.


Gardening
–4
–4
–4
–2
0
0
1
–1
–5
–3

15
–5
–9
–6
3
10
5
9
59
Cultural activities
Pubs/bars/clubs
–6
–4
–2
–4
–5
2
0
4
7

–5
2
3
6
2
5
1
1
–5
53
There is a very clear subgroup of those who participate more in the various
Deviations from average duplications in UK arts and leisure market 2009/10

Go to cinema
–9
–6
–4
–4
–5
1
–1
1

4
–9
–4
4
4
4
10
4
8
1
54
forms of cultural activity, such as arts/crafts, theatre, visiting museums,
Venue

Sport/exercise
–9
–6
–5
–7
–5
–1
0

4
4
–4
2
3
0
6
4
6
7
1
61
galleries and historic places. Participants in one of these cultural activities
Days out
are more likely than expected to participate in each of the other forms
–10
–6
–5
–6
–5
1

0
2
–1
–2
1
–2
–5
–1
6
13
12
7
76
Eat out as shown by the big positive deviations in every cell of this grouping.
A striking feature of the cultural grouping is that these rows of Table 3
–9
–6
–5
–4
–4

2
0
6
4
–3
–1
0
–4
0
9
7
7
1
79
Read (i.e. activities participated in by those people who participate in each of
–6
–4
–3
–3

–1
0
–2
0
–4
0
–3
–2
–5
2
6
8
9
9
76
Shopping
the cultural activities) contain the majority of the positive bold items in
the whole table that are outside of the highlighted groupings for each
–3
–2
–1

–1
2
1
–3
4
1
0
–1
1
0
–6
2
1
2
3
77
Everyday

Listen to music partition. Participants in cultural activities not only duplicate participation
–7
–4

–3
–2
–1
0
–2
2
1
–5
–2
0
2
7
4
3
5
3
84

more with other cultural activities, but also are more likely to engage in
almost all forms of free-time activity. This is confirmation of Peterson’s
Friends/family
–4

–1
–1
–2
1
2
–1
1
1
–2
–1
0
–2
–2
3
3
4
2
91

Watch TV notion of a group of ‘omnivores’ particularly located in the consumers of



0
–1
2
0
0
0
–1
1
2
1
1
–1
2
–5
0
1
0
–1
91

culture. Rather than an elite group eschewing popular pastimes in favour


Sources: authors based on Taking Part Data

of aesthetic pursuits (although a few of those people may exist), we have a


large group engaging with many forms of the Arts and other more populist
Spend time with friends/family

activities. They are no less likely than the rest to watch TV, go shopping
Days out or visits to places

Play a musical instrument


% point deviations from

or to pubs, or to use the internet, and are especially more likely to read,
Visit museums/galleries
Theatre/music concerts
FREE TIME ACTIVITIES:

AVERAGE duplication
Go to pubs/bars/clubs

Play computer games


Eat out at restaurants

Visits to historic sites

garden, eat out, go to the cinema or take days out. Those who play a
Internet/emailing
column average

musical instrument can also be included in this group, with participants


Listen to music

Arts and crafts


Sport/exercise
Go to cinema

in each cultural activity slightly more likely to take part in that minority
Gardening
Shopping
Watch TV

activity, but particularly those who play an instrument being considerably


Read
Table 3

DIY

more likely to participate in cultural activity.


Everyday Venue Home Solo Cultural

12 13
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

Solo activities
that don’t involve going anywhere or doing anything much. Shopping is
This group is labelled solo, as the three activities involve interaction with slightly different, in that it does involve going out, but recreational shopping
computers and instruments rather than other people. The excess duplication doesn’t involve the same specific venue or purpose as the activities in the
between the computer-based activities is not surprising, however it might venue group and the data pattern puts it firmly in this group. What brings
be more of a surprise to find the strong relationship between playing a this group together is that participants in these activities are less likely to
musical instrument and computer-based activity. It turns out that this engage in nearly every other activity compared to the average. This pattern
relationship is a result of the demographic strength that playing an of the most popular items of choice under-duplicating with most other
instrument has in young males, many of whom are, not surprisingly, items is often seen in other contexts (e.g. brand choice), and is typical of
participants in all forms of computer use. The active musicians also tend the ‘natural monopoly’ phenomenon described earlier (Ehrenberg et al.
to listen to music more and take part in sport and cultural activities as 2004). This occurs because the leading item of choice is more likely to
mentioned above, whereas the computer games players go to the pub attract light users of a category. Thus, in this case, those with little free
more, but take part less in reading, days out, gardening and almost all the time (light users) are perhaps more likely to engage in the most popular
cultural activities, i.e. less of the activities that are more likely to attract activity (watching TV, spending time with family) and less likely to engage
an older demographic. in more esoteric pursuits. TV watchers and others in this group still
participate in lots of other activities (around nine others on average), but
less than those participating in each activity in the cultural group, who on
Home-makers
average participate in about 12 other activities.
This little grouping of gardeners and DIY-ers suggests a set of people
who are particularly at home at home. The gardeners in particular are
Demographic profiles
less likely to listen to music, go to the cinema, pub or theatre or use the
computer, which almost certainly also reflects an older demographic bias Demographic profiles of participants in the various activities are discussed
for them. Apart from an affinity with gardening, DIY-ers show little else extensively in the DCMS reports (DCMS 2011). We present here an
that deviates from the average levels of participation in other activity. overview of the main patterns of deviation from an average profile of
users of all activities, in other words showing how different the users
of a particular activity are (or, mostly, are not). Following Kennedy and
Venue
Ehrenberg (2001), the profile of participants in each activity is derived across
This group typically engages more in activities carried out by going somewhere a number of demographics: gender, age, with children or not, education,
specific away from home (e.g. restaurant, cinema, gym). What characterises income and ethnic origin. So, for example, 43% of those watching TV are
the group, is not that its members have a particular extra affinity between male, 9% are 16–24 and 33% are 25–44, 29% have children at home,
these activities, but rather that they show no excess propensity, either for and so on, whereas 58% of those playing a musical instrument are male,
other activities within the group or those outside it. In other words, this 18% are 16–24 and 34% 25–44, 27% are in a household with children
group is made up of people who partake in many different activities without at home. The average profile for each demographic is calculated across
showing any particular tendency that puts them into another group. This is all the activities, and then a deviation for each activity from that average.
another way of being omnivorous, taking part in a broad spread of leisure Finally, the mean absolute deviation (MAD) is calculated for each activity
activities without any particular tendency to ‘overindulge’ in any one. by averaging across the deviations for each level of the demographic,
ignoring the sign. The process described is illustrated in Table 4 for the
gender demographic. So the average profile across the activities is 44%
Everyday
male, 56% female. The TV watching profile differs from that average by
The final group, in the top-left corner of Table 3, is characterised by very –1 for male and +1 for female (i.e. 43% male, 57% female), giving a MAD
popular activities (all with penetrations above 70%), and mostly activities for watch TV by gender of 1 (the average of –1 and 1 ignoring the – sign).

14 15
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

Table 4 Illustrating the profile and MAD calculation: gender profiles Table 5 MAD for six demographics

Deviation from Average MAD Gender Age Children Educate Income Ethnic
Profile column average Everyday Watch TV 1 2 2 2 1 1
Male Female Male Female MAD Time with friends/family 3 2 0 1 1 0
Everyday Watch TV 43 57 –1 1 1 Listen to music 2 1 1 1 1 0
Time with friends/family 41 59 –3 3 3 Shopping 8 2 0 1 2 1
Listen to music 42 58 –2 2 2 Read 7 3 4 1 1 0
Shopping 37 63 –8 8 8 Venue Eat out at restaurants 2 1 2 0 0 0
Read 37 63 –7 7 7 Days out or visits to places 3 1 1 1 0 0
Venue Eat out at restaurants 42 58 –2 2 2 Sport/exercise 4 1 1 1 1 0
Days out or visits to places 42 58 –3 3 3 Go to cinema 3 5 6 1 1 1
Sport/exercise 49 51 4 –4 4 Go to pubs/bars/clubs 6 2 1 1 1 1
Go to cinema 42 58 –3 3 3 Home Gardening 1 6 4 2 1 1
Go to pubs/bars/clubs 50 50 6 –6 6 DIY 14 3 1 1 2 0
Home Gardening 43 57 –1 1 1 Solo Internet/emailing 1 5 7 2 1 1
DIY 59 41 14 –14 14 Play computer games 11 8 9 2 1 1
Solo Internet/emailing 45 55 1 –1 1 Play a musical instrument 14 3 3 3 1 1
Play computer games 55 45 11 –11 11 Cultural Theatre/music concerts 5 2 3 1 1 1
Play a musical instrument 58 42 14 –14 14 Visit museums/galleries 1 3 2 2 1 0
Cultural Theatre/music concerts 40 60 –5 5 5 Arts and crafts 15 2 3 1 2 1
Visit museums/galleries 43 57 –1 1 1 Average MAD 6 3 3 1 1 1
Arts and crafts 29 71 –15 15 15
Source: authors based on Taking Part Data
Average 44 56 0 0 6
Source: authors based on Taking Part Data
MADs for age of respondent and presence of children are also small, but
A quick glance up and down the MAD and adjacent columns in Table 4 where there are deviations they tend to go together for both demographics.
shows that there are some activities with big gender biases relative to the Thus, there is a greater tendency for younger respondents and those with
overall average of 44% male, 56% female: towards males for DIY, playing a child in the household to go to the cinema and use computers, and a
computer games and playing a musical instrument, and to females for greater tendency for older respondents and those without children to read
shopping, reading, and arts and crafts. There are smaller biases to males and garden.
for sport and going to pubs, and to females for going to the theatre and
concerts. Table 5 shows the MAD results from the same analysis carried
Limitations
out for the other demographics, i.e. the equivalent to the final column in
Table 4. The DCMS study has a wealth of more detailed information that we have
For three of the demographics – level of education, income and ethnic not used – for example, breaking out individual activities such as types of
origin – the small values for MAD (2 or under) throughout Table 5 mean sports or crafts within the broader leisure categories reported in this paper,
there are few deviations from the average profile for any activity. An whether participation was active or as a spectator, how many times and
exception (not seen in Table 5 but in the detailed breakdown) is that, for how much it was enjoyed. Further work could include attempting to fit
the highest level of education (degree or higher), there is a slightly greater an NBD-Dirichlet model of choice to activities (as, for example, in Hand
tendency to play an instrument or go to museums and art galleries, and 2011), which might endorse the similarities between leisure choices and
a lesser tendency to watch TV and spend time with family. Many of the those made in other markets.

16 17
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

Conclusions and implications


This being the case, most people will necessarily choose each activity
This article has investigated the main structure of competition for users occasionally, choosing some more regularly than others, of course, but
between free-time activities, in particular to examine whether and how many activities sometimes. This ties directly to the finding that the most
much activities group together in their user base, and what characterises important factor in creating and maintaining a successful brand or product
such groups. is the number of people using it, not how intensely they use it or what
To sum up, there are partitions in the leisure market where activities else they use. Intense users are few and atypical of the wider user base. It
share and therefore compete somewhat more for users with other similar is not that they do not matter, but that their incidence tends to follow in
types of activity: especially one that groups the users of cultural-based line with how many users there are overall – a pattern known as ‘double
activities. Nevertheless the dominant pattern is that the main determining jeopardy’ (McPhee 1963).
factor in any group’s incidence of using another free-time activity is the A consequence is that growth, where it occurs, almost always comes
popularity of the other activity (i.e. the number who take part at all). from attracting usage across the potential user base including light and
For example, although theatre-goers may be 10% more likely than non-users (penetration growth), and not simply through getting existing
the average activity user to be involved in arts and craft, far more or heavier customers to consume more (Sharp 2010). This has a major
theatre-goers (58%) go to the pub than do arts and craft (31%). Most positive implication because the traditional marketing for culturally
people participate in multiple activities, in line with Peterson’s view of based activities has focused on a small, narrowly defined target of
the omnivorous consumer (Peterson 2005), which is in turn akin to culturally inclined consumers who are more likely to respond positively
Ehrenberg’s description of polygamous brand users (Ehrenberg et al. to a marketing intervention. However, this study makes clear that growth
2004). The data show that people mainly do not migrate to cultural for arts organisations is more likely to come to those that promote their
activity at the expense of popular culture – rather they adopt cultural activities regularly and widely, and seek new audiences (Rentschler 2002).
activities in addition. Moreover, any culturally based activity that aims to increase its level of
Given much multiple activity participation and minor partitioning of participation should expect to attract participants from all other leisure
activities, there are only a few differences in the demographic profiles of activities, not just from other cultural ones, albeit there would be a slightly
participants in each activity. Gender in particular, age and the presence more than proportional draw from them.
of children are more likely to influence participation in certain leisure Another lesson from other consumer markets is that, since only a
activities, rather than education or income. small part of the user base is heavily involved, the primary function of
The patterns are similar to those found in many other circumstances marketing is to constantly remind its base of mostly occasional users
where consumers make repetitive choices between alternatives, such as to continue taking part. After all, with occasional activities, months or
in consumer-packaged goods. The notion of distinct segmentation, where even years may elapse between usage occasions. Thus, not only growth
the stereotypical consumer of culture is a highbrow person who would requires constantly reminding or nudging of consumers, but so does
not wish to be associated with common leisure pursuits, and vice versa, maintaining the existing level of involvement. This in turn leads to the
is contradicted by the data. Instead, the slight segmentation for these need to establish reach as the main goal for any form of marketing
activities occurs in the broader context of wide participation across the communication (Sharp 2010).
range of leisure activities. This is the same as is found in many other Leisure activities are different in many ways from other consumer
consumer markets – slight segmentation for some highly targeted, but goods – they can be more involving, more loaded with social connotation
generally small, brands. We expect, then, that the lessons learned in other and require greater effort for participation. However, despite these
consumer markets will apply similarly in the markets for cultural and differences, the underlying behaviour exhibited by people choosing to
other leisure products. participate in leisure activities is very much like those of other consumer
A key general finding is that people engage in a large number of markets. Thus, the choices that drive the spending of money, of time and
leisure activities – on average nearly a dozen separate activities. In other even of social capital all adhere to the same laws of consumer choice
words, they choose activities across a very wide portfolio of preferences. behaviour.

18 19
The competitive landscape for leisure International Journal of Market Research Vol. 57 Issue 2

References Peterson, R.A. (2005) Problems in comparative research: the example of omnivorousness.
Poetics, 33, 5–6, pp. 257–282.
Agostini, J. M. (1961) How to Estimate Unduplicated Audiences. Journal of Advertising Rentschler, R. (2002) The Entrepreneurial Arts Leader: Cultural Policy, Change and
Research, 1, 11–14. Reinvention. St Lucia: University of Queensland Press.
Bennett, D.R. & Graham, C. (2010) Is loyalty driving growth for the brand in front? Journal Scriven, J.A. & Danenberg, N. (2010) Understanding how brands compete: a guide to
of Strategic Marketing, 18, 7, pp. 573–585. duplication of purchase analysis. Report 53 for Corporate Supporters, Ehrenberg Bass
Bourdieu, P. (1984) Distinction: A Social Critique of the Judgment of Taste (translation, Institute, University of South Australia.
originally published 1979). Cambridge, MA: Harvard University Press. Sharp, B. (2010) How Brands Grow: What Marketers Don’t Know. Melbourne: Oxford
Colbert, F. (1997) Changes in marketing environment and their impact on cultural policy. University Press Australia and New Zealand.
Journal of Arts Management, Law, and Society, 27, 3, pp. 177–186. Stebbins, R.A. (2007) Serious Leisure: A Perspective for Our Time. New Brunswick, NJ:
Dawes, J.G. (2008) Regularities in buyer behaviour and brand performance – the case of Transaction Publishers.
Australian beer. Journal of Brand Management, 15, 3, pp. 198–208. Uncles, M.D., Ehrenberg, A.S.C., & Hammond, K.A. (1995) Patterns of Buyer Behavior:
Department for Culture, Media and Sport (DCMS) (2010) Taking Part: the National Survey Regularities, Models, and Extensions. Marketing Science, 14(3), G61–G70.
of Culture, Leisure and Sport, 2009–2010. Adult and Child Data (computer file), UK Data
Archive (distributor), October. SN: 6579. Colchester, Essex: DCMS.
Department for Culture, Media and Sport (DCMS) (2011) Taking Part: the National Survey of
Culture, Leisure and Sport. Available online at: www.culture.gov.uk/what_we_do/research_ About the authors
and_statistics/7184.aspx (accessed 14 August 2011).
Ehrenberg, A.S.C. (1982, reprinted 1994) A Primer in Data Reduction. Chichester & New John Scriven is Senior Research Associate at the Ehrenberg-Bass Institute,
York: Wiley. University of South Australia. His research interests major on the nature
Ehrenberg, A.S.C. (1988) Repeat Buying: Facts Theory and Applications, 2nd edn. London: of brand loyalty and competition, and the effect of marketing inputs,
Edward Arnold; New York: Oxford University Press.
Ehrenberg, A.S.C. & Goodhardt, G.J. (1969) Practical applications of the duplication of
particularly price and advertising. He has a particular interest in marketing
viewing law. Journal of the Market Research Society, 11, January, pp. 6–14. metrics and the development of numeracy. Prior to 20 years an academic,
Ehrenberg, A.S.C., Uncles, M. & Goodhardt, G.J. (2004) Understanding brand performance he had over 20 years commercial experience with three major international
measures: using Dirichlet benchmarks. Journal of Business Research, 57, 12, pp. 1307– corporations.
1325.
Gilmore, A. & Rentschler, R. (2002) Changes in museum management: a custodial or Diana C. Pérez-Bustamante Yábar is Associate Professor at Rey Juan
marketing emphasis? Journal of Management Development,21, 9/10, pp. 745–760. Carlos University in Madrid. She holds a Ph.D. with Summa Cum Laude,
Glenn, N.D. (1969) Social Stratification. New York: Wiley. and an MBA in marketing from ESIC Business & Marketing School.
Goodhardt, G.J. (1966) The constant in duplicated television viewing. Nature, 212, December,
p. 1616.
She collaborates and participates in research international teams, such
Goodhardt, G.J., Ehrenberg, A.S.C. & Collins, M.A. (1987) The Television Audience – as London South Bank University, the Ehrenberg Centre for Research in
Patterns of Viewing: An Update, 2nd edn. Hampshire: Gower Publishing Co. Ltd. Marketing, London College of Fashion, University of Arts. Her research
Hand, C. (2011) Do arts audiences act like consumers? Managing Leisure, 16, 2, pp. 88–97. interests include the nature of consumers’ behaviour and brand loyalty
Hand, C. & Singh J. (2014) Segmenting the betting market in England. International Journal
of Market Research, 56, 1, 111–127. focusing on cultural and creative industries, fashion and arts.
Kennedy, R. & Ehrenberg A.S.C. (2001) There is no brand segmentation. Marketing Insights, Maria Clemente is a freelance Visiting Senior Researcher at London
Marketing Research, 13, 1, pp. 4–7. South Bank University. She has worked and studied in the UK and Brazil
Lees, G. & Wright, M.J.R. (2013) Does the duplication of viewing law apply to radio
listening? European Journal of Marketing, 47, 3/4, pp. 674–685.
and has 11 years practical experience in social/marketing research in
Lynn, M. (2013) The target market misapprehension: lessons from restaurant duplication quantitative research methods. Her research interests are wide-ranging,
of purchase data. Cornell Hospitality Report, 13, 3, February. New York: Center for but the focus of her work in recent years has been on analysing
Hospitality Research, Cornell University. consumer behaviour of FMCG brands. She is interested in new products,
McLean, D.D. & Hurd, A.R. (2012) Kraus’ Recreation and Leisure in Modern Society, 9th
edn. Sudbury, MA: Jones and Bartlett. brand-performance measurement, segmentation and positioning, and the
McPhee, W.N. (1963) Formal Theories of Mass Behaviour. Glencoe, NY: Free Press. application of multivariate analysis and marketing research techniques.
Mencarelli, R., Marteaux, S. & Pulh, M. (2010) Museums, consumers, and on-site Dag Bennett is a Reader in Marketing at London South Bank University,
experiences. Marketing Intelligence & Planning,28, 3, pp. 330–348.
OECD (2009) Society at a Glance 2009. OECD Social Indicators. Paris: OECD.
Director of the Ehrenberg Centre for Research in Marketing and Senior
Peterson, R.A. (1992) Understanding audience segmentation: from elite and mass to omnivore Adjunct with the Ehrenberg-Bass Institute, University of South Australia.
and univore. Poetics, 21, 4, pp. 243–258. His research focuses on brand performance metrics using a two-purchase

20 21
The competitive landscape for leisure

analysis technique. Prior to becoming an academic, he had 15 years of


commercial management experience with Procter & Gamble and other
companies
Address correspondence to: John Scriven, Ehrenberg-Bass Institute for
Marketing Science 7, Milkingpen Lane, Basingstoke RG24 7DD.
Email: john.scriven@marketingscience.info

22

View publication stats

You might also like