Buying Brands at Both Regular Price and on Promotion over Time

Abstract

We analyse the purchasing of brands at both regular and promotional price over time. The goal is to better understand the extent of consumer deal-proneness. Our analysis shows most consumers buy brands on promotion at least some of the time, and the tendency to buy on promotion relates mostly to how much promotion is available in a category, suggesting little innate deal-proneness. The extent of promotion can be so high that as many as half of all brand buyers buy the brand solely when it is on promotion. However, this amount of on-deal buying is only very slightly higher than would be expected given the amount of promotion available. We find few buyers buy only on promotion. Promotion buyers of a particular brand also buy other brands on and off promotion more or less in line with the market share those other brands have at regular and promotional price. The three main implications are: (1) brand loyalty is still an important aspect of purchase, (2) a brand's normal-price buyers are a major source of its volume from price promotions, and (3) there is only a small effect of deal-proneness on promotion buying over and above that of promotion prevalence in a category.

Full text

1
Buying Brands at both Regular Price and on Promotion Over Time
John Scriven
Maria Clemente
John Dawes* corresponding author
Giang Trinh
Byron Sharp
Ehrenberg-Bass Institute for Marketing Science
Contact: John.Dawes@marketingscience.info
Published in Australasian Marketing Journal, Vol 25, 4, 2017

2
Buying Brands at both Regular Price and on Promotion Over Time
Abstract
We analyse the purchasing of brands at both regular and promotional price over time. The goal
is to better understand the extent of consumer deal-proneness. Our analysis shows most
consumers buy brands on promotion at least some of the time, and the tendency to buy on
promotion relates mostly to how much promotion is available in a category, suggesting little
innate deal-proneness. The extent of promotion can be so high that as many as half of all brand
buyers buy the brand solely when it is on promotion. However, this amount of on-deal buying
is only very slightly higher than would be expected given the amount of promotion available.
We find few buyers buy only on promotion. Promotion buyers of a particular brand also buy
other brands on and off promotion more or less in line with the market share those other brands
have at regular and promotional price. The three main implications are: (1) brand loyalty is
still an important aspect of purchase, (2) a brand’s normal-price buyers are a major source of
its volume from price promotions, (3) there is only a small effect of deal-proneness on
promotion buying over and above that of promotion prevalence in a category.
Keywords: price promotion; deal-prone buyer; brand loyalty

3
Buying Brands at both Regular Price and on Promotion Over Time
1. Introduction
Price Promotions1 are a very expensive activity for consumer packaged goods marketers.
Between 50% to 60% of CPG firms’ marketing budgets is reportedly spent on price
promotions (Bolton et al., 2010, Nielsen, 2009). Marketers, finance directors, as well as
academics worry that costly price promotions have no positive impact on the brand long-term
(Jones, 1990, Nijs et al., 2001).
Extensive literature on the topic generally agrees that price promotion produces a short-term
spike in sales that then return to normal when the promotion finishes (e.g. Dawes, 2004,
Drechsler et al., 2017, Totten and Block, 1994). Furthermore, some studies have suggested a
possibility that price promotions have undue appeal to only a portion of buyers – a promotion-
buying or deal-prone buyer segment (Lichtenstein et al., 1995, Webster, 1965). A deal-prone
buyer is defined as having a high tendency to buy brands on promotion, over and above the
extent of price promotion in the category (Webster, 1965). Deal-proneness also implies a
tendency to switch between brands to take advantage of deals (Dodson et al., 1978).
Aside from marketer’s concern about expense and lack of positive long-term impact, there is
also worry that price promotions may train consumers to become deal prone, moving their
loyalty from brands to the deal itself (e.g. Mela et al., 1997). That concern provides the central
motivation for this research. The study aims to understand how the overall prevalence of deals
or price promotion in a category is related to how many households buy brands only on
promotion / deal, only at normal price, or both over a one-year time period. The findings will
also help to clarify the influence of brand loyalty in promotion-intense categories.
2. Background and Literature Review
The concept of the deal-prone buyer has a long history. Deal-proneness has been generally
investigated as a consumer trait that is linked to demographics, resource constraints (e.g. ability
to travel), and category usage rate (Blattberg et al., 1978). Other studies have sought to extend
the list of traits to psychographic and shopping-related variables to characterize more and less
deal-prone consumer segments (Martînez and Montaner, 2006, Palazon and Delgado-Ballester,
2011). But Webster (1965) identified early on that deal proneness is complex to measure,
because it is confounded by the amount of deals offered by brands. Moreover, deal-proneness
will be influenced by retailer decisions. That is, if consumers tend to buy at retailers that
constantly run promotions, a large proportion of their purchases will be made on-promotion.
But they may not be necessarily deal-prone. Indeed, Pechtl (2004) found nearly 50% of
consumers were not deal-prone and that deal-proneness had only a weak link with shopping at
either an Every Day Low Price (EDLP) or Hi-Low price retailer. Furthermore, the study
showed self-identified deal-prone shoppers actually bought less than 20% of their grocery
items on deal.
Another stream of research has examined the extent to which households might become more
deal-prone or promotion-sensitive over time. To this end, a number of studies have sought to
1 We use the terms ‘price promotion’ and ‘deal’ interchangeably, meaning a temporary offer involving a lower
price or extra value to consumers.

4
identify whether a household’s purchase history alters its current purchase behavior. For
example, Hardie, Fader and Johnson (1993) found evidence that past purchases for a brand on-
deal made households slightly less likely to buy the brand again at regular price. Mela, Jedidi
and Bowman (1998) found increased incidence of promotions in a category resulted in slightly
less frequent purchases, with slightly more quantity bought each time. The theoretical
explanation for these phenomena invoke the concept of a reference price (Mazumdar et al.,
2005), which means consumers remember and are influenced by previous prices paid. If a
previous purchase was made at a reduced price, it is believed that the consumer’s reference
price for the brand is lowered. Consumers are thought to be resistant to pay more than their
reference price, in line with loss aversion (Klapper et al., 2005, Bell and Lattin, 2000).
Kalyanaram and Winer (1995) support this view, concluding “reference prices have a
consistent and significant impact on consumer demand” (1995, G 163). If a shopper’s
reference price is lowered, logically this should induce them to seek deals, hence become more
deal-prone.
Therefore, there is some evidence that deal-proneness is not necessarily just a fixed trait arising
from household variables (as per Blattberg et al., 1978), but something that may be affected by
brands’ and retailers’ promotion activity. The increasing incidence (Bogomolova et al., 2015),
and prevalence of promotions for brands in packaged goods markets (Nielsen, 2009) should
arguably make for a large, and growing, deal-prone segment. However - is there really a large
proportion of shoppers who are quite deal-prone, that is, purposively selecting only brands that
are price-promoted at the time? If there is, it implies brand loyalty should be declining, with
more consumers readily swapping between any brand to take advantage of price deals. Yet
multiple studies have found considerable stability in brand loyalty over time (Johnson, 1984,
Dawes et al., 2015). Furthermore, whilst it seems reasonable that repetitive promotions can
train consumers to buy on deal (Mela et al., 1997), several studies show little effect of leading
brands regularly running temporary deals or promotions. Ehrenberg, Hammond and
Goodhardt (1994) found a high level of stability in an analysis of repeat-purchasing before and
after promotions. Dekimpe, Hanssens and Silva-Risso (1998) found brand sales to be
predominantly stationary in the medium term in markets characterized by frequent promotion
activity.
The literature therefore portrays a mixed picture on the issue of how promotions might
exacerbate deal (or promotion) proneness. Some evidence exists that price promotions should
erode reference prices, heightening deal-proneness. However, several studies show little effect
in terms of brand sales or repeat-purchase. To address this somewhat confusing picture, the
paper presents an empirical analysis of sixteen categories of packaged consumer goods with
differing levels of price promotion, examining how consumers purchase brands at both
promotion and regular price. The intended contribution is to establish if we can observe any
empirical generalisations across a diverse set of packaged consumer goods categories,
concerning how many households buy categories and brands on promotion, compared with the
overall amount of promotion occurring. The time period of the analyses is one year. The study
then goes on to examine how buyers of brands at regular and promotional price buy other
brands at regular and promotional price, to establish how much, if any, partitioning there is
between regular and promotional price buying. This cross-purchase analysis in turn provides a
measure of how much consumers seek out promotions, compared with simply taking
advantage of promotions because they are available.
2. Method

5
The data consisted of purchasing records in two different time periods for Fast Moving
Consumer Goods (FMCG) categories kindly provided by Kantar Worldpanel from its UK
consumer panel. The panel comprises a demographically balanced sample of over 15,000
households, who scan their grocery purchases (Kantar, 2015). We use eleven product
categories from 2007, and six from 2014. One category, Fabric Care, appears in both time
periods. We also use data from different time periods in an effort to produce generalisable
results across categories and time. We use categories that vary greatly in average purchase
frequency, as shown in Table 1. Figures for purchase frequency were not able to be extracted
for three categories (the data is provided via a database-like format that occasionally precludes
certain calculations).
Our choice of categories was based on data availability. The categories comprise only a
sample of all consumer goods, but include food, beverage, cleaning, and personal care. Our
analysis is of purchase-based metrics, so promotion incidence is measured by percentage of
total sales made on deal, not weeks on deal nor depth of discount. Purchases were classified as
either being made at regular price, or on price-related promotion (cut price, extra volume free,
buy one get one free etc.). Share of purchases and penetration (% of households buying) were
calculated for regular price and promotional price for each category and the five largest brands
in each category. Penetrations were broken down into those who only bought at regular price,
those who only bought at promotional price and those who bought at both in the year. Hence
penetration at promotional price is the sum of penetration at promotional price only, and at
both. We also calculated the proportion of shoppers who only bought at regular or promotional
price that were one-time category buyers - to identify the extent that light category buying is
related to promotion-only or regular-only buying.
Next, we conducted purchase duplication analysis (Tanusondjaja et al., 2016, Dawes, 2016) of
the brands in each category, with each brand split into regular and promotion price purchases.
The duplication analysis method calculates the proportion of those buying brand A that also
buy other brands B, C, D and so on in a time period. The method is versatile, having been
principally applied to brands (Ehrenberg, 2000), but also in other contexts such as the cross-
purchasing of retailers (Dawes and Nenycz-Thiel, 2014), and buying of wine across different
price tiers (Romaniuk and Dawes, 2005).
3. Findings
There is a great diversity in the incidence of price promotion across the categories, from 53%
of purchase occasions in Fabric Care (2014), 51% for Pizza (2007) and Instant Coffee (2014),
to only 7% in condiments (Table 1, column 2). Biscuits appear to be quite different to the
other categories with very low incidence of buying only promotion, or only at regular price –
this is likely due to its very high average purchase frequency of 28 occasions per year.
Apart from the categories with low incidence of promotion – Razor Blades and Condiments &
Spices, the majority of category buyers buy on promotion at some time during the year. We
see that the figures are generally above 50% in the ‘Promo at all’ column. However, very few
(less than 10% of category buyers on average, and at most 19%, see ‘Only at Promo’ column)
buy the category only when via a price promotion, even in the very heavily promoted
categories. Almost everybody – around 90% who buy a category – buy a brand at regular price
at some time. This is quite apparent in the ‘Regular at all’ column with averages of 93% and
86%. Therefore, it appears there is only a fairly small proportion of shoppers who will only

6
buy a category when they can take advantage of a price promotion. There is never more than
around 20% of buyers who only buy the category on promotion, and the averages across the
categories for the two years are 7% and 9%.
The question arises as to how much of the incidence of buying only on promotion (or only at
regular price) in a year pertains to the prevalence of very light buyers, who buy the category
only once in the year. In Table 1 we include the proportion of households that bought only at
regular price, or only on promotion – that are also one-time category buyers. Overall, this
proportion is 39%. This indicates that a reasonable proportion (but a minority) of promotion-
only buying occurs simply because the shopper bought the category only once. Also, the
overall extent of promotions in the category will still influence whether a one-time buyer buys
on promotion or not, when they do happen to buy the category.
We can also make some observations about how increases in overall promotion prevalence
lead to increases in only buying on deal. In 2007, Fabric Care had 29% of all purchases on
promotion and 5% of households bought only on promotion. In 2014 the proportion of all
purchases on promotion rose to 53%, and households buying only on promotion rose markedly
to 18%.

7
Table 1 Category buying at regular and promotion price in a year
52 weeks 2007
%
purchases
on
Promotion*
Percent of Category buyers buying …
Category
Purchase
Frequency
% of only
regular or promo
buyers, that buy
category only
once
Categories
Only
at
regular
Only
at
promo
Both
Promo
at all
Regular
at all
Chilled &
Frozen Pizza
51
17
11
72
83
89
7
53
Toilet Tissue
35
18
8
74
82
92
11
18
Yoghurt Drinks
35
33
17
50
67
83
na
na
Biscuits
32
2
0
98
98
100
28
20
Fromage Frais
29
38
10
51
62
90
9
49
Fabric Care
29
28
5
67
72
95
8
61
Nappies
28
31
7
62
69
93
9
48
Baked Beans
17
40
4
56
60
96
10
16
Air Fresheners
16
58
4
38
42
96
na
na
Razor Blades
9
84
5
10
16
95
na
na
Condiments &
Spices
7
72
1
27
28
99
5
22
Average 2007
26
38
7
55
62
93
11
36
52 weeks 2014
Fabric Care
53
19
18
62
81
82
6
30
Instant Coffee
51
19
19
62
81
81
8
38
Dental Care
47
22
15
63
78
86
5
35
Ice-cream
45
14
8
78
86
92
10
39
Beer
43
25
14
60
75
86
8
62
Skin Care
35
37
11
52
63
89
4
57
Average 2014
45
23
14
63
77
86
9
44
Average
overall
33
33
9
58
67
91
9
39
Underlying data from Kantar Worldpanel. * note these % purchases on promotion figures are weighted average for category and may differ
slightly in subsequent tables, which show simple unweighted averages.
3.1 Results for specific brands
What do these metrics look like for brands within these categories? Does the amount of
promotion in a category reflect the amount of promotion that goes on for each brand? For
example, are all brands in heavily promoted categories also heavily promoted, or do some
brands resist the tactic even though it is the norm? Do some brands promote heavily even if
the category is not that heavily promoted? In general we find that brands follow their category
norm: in heavily promoted categories all brands offer a lot of deals. We show detailed results
for several of the categories to highlight specific points in detail. We then combine all the
categories and use them in Table 8 to convey generalized findings and take-outs.
Table 2 shows the top five brands in Chilled & Frozen Pizza, the most heavily promoted
category in our 2007 data (51% of all purchase occasions bought on promotion, 50% among
the top 5 brands). All the brands are heavily promoted, with the three leaders each having
59%+ of purchase occasions made on promotion.

8
Table 2 Brand buying at regular and promotion price in a year: Pizza
Brand*
Market
Share %
% on
Promotion
Percent of brand buyers buying brand
Only at
regular
Only at
promo
Both
Promo
at all
Regular
at all
Goodfellas
15
67
20
44
37
80
56
Chicago Town
12
65
19
44
37
81
56
Dr Oetker Ristorante
4
59
24
50
26
76
50
Pizza Express
1
31
58
23
19
42
77
M+S
1
27
60
21
19
40
79
Average
7
50
36
36
28
64
64
Underlying data from Kantar Worldpanel *there are a plethora of other small brands in this category not reported.
About two-thirds (64%, range = 80% to 40%) of each of these brands’ buyers made a brand
purchase on promotion (next to last column), and 36% on average only bought the brand when
on promotion. Even the small share premium brands Pizza Express and M&S have about 30%
of occasions on promotion and around 40% of their buyers buy sometime on deal, although the
proportion of ‘Only at promo’ buyers for these two brands is lower at approximately 20%.
Table 2 shows there is a high incidence of promotion-only buying for these leading Pizza
brands, coupled with only about half their buyers buying at regular price. This suggests a
strong propensity to switch between these leading brands for a deal. If the promotion-only
buying figures were lower, it would indicate that buyers of any particular brand alternated
between buying that same brand on and off promotion, rather than switching between the
leading brands. In part this switching may be due to the high frequency of deals (so that
probably at least one of the major brands can be found on deal at any time). A lower incidence
of overall promotion would likely have resulted in consumers merely taking a deal when
available (rather than switching for a deal), which we would see reflected in a higher incidence
of buying at both regular and promotional price. Interestingly, the figures for ‘Both’ in Table 2
decline in line with brand size, which suggests a Double-Jeopardy effect – small brands are
bought a bit less often, so will have fewer buyers buying them at normal price and on
promotion.
The Yoghurt Drinks category is somewhat similar, as Table 3 shows. The main brands,
Danone and Muller with 39% and 26% market share respectively have 37% and 55% of sales
respectively on promotion. 65% of Danone buyers buy it sometime on promotion and 24%
only on promotion. For Muller, it is 74% buying it sometime on promotion, and 39% only on
promotion. An interesting oddity in this category is the brand Yakult, the founder of the
category but now a minor brand with 9% share. It does almost no price promotion at all, and
hence has very few promotion buyers and almost none who only bought it on promotion.
Table 3 Brand buying at regular and promotion price in a year: Yoghurt Drinks
Brand
Market
Share
%
% on
Promotion
Percent of brand buyers buying brand
Only at
regular
Only at
promo
Both
Promo
at all
Regular
at all
Munch Bunch
5
55
28
36
36
72
64
Muller
26
55
26
39
35
74
61
Danone
39
37
35
24
41
65
76
Benecol
5
26
45
16
39
55
84
Yakult
9
1
96
2
2
4
98
Average
17
35
46
23
31
54
77

9
Underlying data from Kantar Worldpanel
In contrast, Table 4 shows the same metrics for brands in Condiments & Spices, where there is
little promotion on any brand. Here every brand is closer to Yakult in drinking yoghurt - for
most brands in this category, fewer than 15% of their buyers buy it on promotion at all (% on
Promotion column), and hardly any buy a brand only when it is on promotion: an average of
5%.
Table 4 Brand buying at regular and promotion price in a year: Condiments & Spices
Brand
Market
Share
%
% on
Promotion
Percent of brand buyers buying brand
Only at
regular
Only at
promo
Both
Promo
at all
Regular
at all
Saxa
6
13
82
10
8
18
90
Colmans Mustard
8
10
86
6
8
14
94
Lo Salt
2
6
91
6
4
9
94
McCormic
4
2
97
2
2
3
98
Schwartz
15
2
96
2
2
4
98
Average
7
7
90
5
5
10
95
Underlying data from Kantar Worldpanel
We show one more example table of this type, using a category from our 2014 data, Ice-cream,
which had a very large proportion of occasions purchased on deal (80%). We see the brand
with the largest amount of promotion, Ben+Jerry’s sold 85% on promotion, meaning that few
of its buyers were able to buy it only at regular price (9%) and 73% bought it only on
promotion. However, 27% of its buyers still bought it at regular price at least once. By
comparison, Viennetta, the brand with the least (comparatively) amount of promotion sales, but
still high at 70%, had a much larger proportion buying it only at regular price (17%) and even
with 70% in total sold on promotion, 37% of its buyers managed to buy it at regular price at
least once.
Table 5 Brand buying at regular and promotion price in a year: Ice-cream
Brand
Market
Share
%
% on
Promotion
Percent of brand buyers buying brand
Only at
regular
Only at
promo
Both
Promo
at all
Regular
at all
Ben+Jerry’s
2
85
9
73
18
91
27
Magnum
8
83
8
69
23
92
31
Cornetto
4
82
11
74
15
89
26
Carte D’Or
3
81
12
72
16
88
28
Viennetta
2
70
17
63
20
83
37
Average
4
80
11
70
18
89
30
The previous four tables show that the more of a brand’s sales that are on promotion, the more
its buyers buy on promotion and there are more who have only bought on promotion. It is
unsurprising that these metrics are related, logically if a brand sells 100% on deal then also
100% of buyers of it will buy on deal and no one at regular price. While if very few sales are
on deal (like the rarely promoted Yakult) then almost no one will buy only on deal. But what
is surprising is the strength of the relationship, it is near perfect, and perfectly linear, i.e. if the
proportion a brand sells on deal is double a rival brand then the percentage of its buyer who
buy on deal is also double. Figures 1 and 2 illustrate the strong relationship (R2 = .88, .89)
across between the percentage of all occasions on which a brand is bought on deal, and the
percentage of buyers who only buy the brand when on deal. This pattern is equally apparent in

10
2007 and in 2014.
These results strongly suggest that there is not an appreciable segment of deal-prone
consumers. If such a segment existed then the relationship would not be linear. For example,
if a brand’s customer base consisted of ‘loyals’ and ‘switchers,’ as is sometimes portrayed in
the academic literature (e.g. Rossiter and Bellman, 2005), we should expect something of a
floor for the percentage of buyers buying on deal. If half of a brand’s buyers were switchers,
then even if a small proportion of a brand’s sales were on deal in a period around half of its
buyers should have bought on deal, i.e. the switchers rush in when the brand is on deal, even if
they don’t contribute a great deal of overall sales (being swamped by the more regular buying
of the ‘loyals’) they make up a large proportion of the buyer base and most of those buying on
deal. The linear relationship between these metrics does not fit with the theory of a brand’s
customer base being made up of very different segments of ‘loyals’ and ‘switchers’. The
customer base is then much better described as being one of polygamously loyal buyers with
differing weights of purchase frequency (as described in Sharp et al., 2012).
Figure 1 & 2 Relationship between overall promotion incidence and buying brand only
on promotion
Underlying data from Kantar Worldpanel
3.2 Duplication Analysis
In order to further explore deal proneness, the last part of the analysis looks at how buyers
duplicate their purchasing between brands on and off promotion (i.e., buy both on & off
promotion in the time period). The key question is, are buyers of one brand on promotion
much more likely to buy other brands on promotion and less likely to buy brands at regular
price? Again we illustrate with the Pizza market in Table 6 before summarizing with averages
across all the categories.

11
Table 6 Duplication of buying at regular and promotion price in a year: Pizza brands
Pizza
Buyers of:
% who also buy:
Regular price
On Promotion
Regular price
Good
fellas
Chicago
Town
Dr
Oetker
Pizza
Express
M&S
Good
fellas
Chicago
Town
Dr
Oetker
Pizza
Express
M&S
Goodfellas
-
33
11
4
3
65
42
15
2
1
Chicago Town
40
-
8
4
3
51
67
12
2
1
Dr Oetker
41
26
-
7
5
47
35
51
3
2
Pizza Express
22
17
10
14
29
19
10
26
6
M&S
18
12
8
15
21
15
7
6
22
Avg. (ex diagonal)
30
22
9
8
6
37
28
11
3
3
On Promotion
Goodfellas
45
30
8
4
3
-
44
17
3
2
Chicago Town
34
46
8
3
2
52
-
14
2
2
Dr Oetker
39
26
33
5
3
61
43
-
3
2
Pizza Express
23
19
7
48
10
34
23
12
-
4
M&S
16
14
6
13
45
27
24
10
5
-
Avg. (ex diagonal)
28
22
7
6
4
43
34
13
3
2
Penetration
20
16
5
4
3
28
24
8
2
2
Underlying data from Kantar Worldpanel
The top half of Table 6 shows the percentage of buyers of each brand at regular price (the
rows) who also bought each other brand at regular price (the first five columns) and each brand
(including the brand in question) at promotional price. The bottom half of the table gives the
equivalent figures among those who bought each brand on promotion. The bottom row of the
table shows the penetration, i.e. percent buying in the population, for each brand at regular or
promotion price. The main pattern in the duplication table is in line with the Duplication of
Purchase Law, that buyers of X who also buy Y is proportional to the size (penetration) of Y
(Ehrenberg and Goodhardt, 1968, Ehrenberg and Goodhardt, 1969, Goodhardt and Ehrenberg,
1969, Tanusondjaja et al., 2016). Thus the numbers in the columns are similar to each other,
but differ from column to column (descending from left to right within each quadrant, see
averages rows). There is a little variation within a column due to some partitioning between the
premium brands Pizza Express and M&S and the others. There is then one obvious big
deviation, that the diagonals of the NE and SW quadrants (highlighted in bold) representing the
buyers of a brand at regular who also buy that brand on promotion and vice versa, are much
higher than the rest of their column. This tells us that:
(1) people attracted to buy a brand on promotion are highly likely to also buy it at regular price,
and conversely,
(2), those who buy a brand at regular price are a major source of custom when the brand is
promoted.
This is clear evidence of brand loyalty, something widely observed across markets, and a
natural part of human behaviour (see Sharp, 2010, Sharp, 2013).
There is a further small, but consistent deviation from the Duplication Law, which is best seen
in Table 7, which takes the averages plus the diagonals from Table 6. In Table 7, row 1 is
always slightly higher than row 2, summarized by the averages of 15 to 13. This says that
among buyers of a brand at regular price, the average sharing with each competitor is 15%.
Then, among buyers of a brand at promotion price, average sharing is similar, at 13%.
Conversely, if we look at rows three and four, row four is generally higher than row three
(average 19 v 16). Thus the buyers of a brand on promotion are a little more likely to buy

12
another brand on promotion than buyers at regular price are to buy that brand on promotion.
These small differences represent the amount of price proneness over and above what we
might expect simply from the availability of price promotion. That is, promotion buyers are a
little more likely to buy another brand on promotion than regular buyers, and regular buyers
are a little more likely to buy another brand at regular price than promotion buyers are. In rows
5 and 6 we see the proportions of regular-price buyers of a brand who buy the same brand at
promo price; and vice versa. These figures are 46 and 43 respectively. The small 3 point
difference here is due to lower promotion incidence for Pizza Express and M&S: if these
brands have fewer promotions, then there is less opportunity for regular buyers to also buy the
brand on promotion. Overall, the figures of 46 and 43 (average 44) can be compared to the
other figures (15 and 13, 16 and 19 = average 15) to indicate the extent to which buyers buy
the same brand on and off promotion, compared to buying any other brand on and off
promotion: about 44/15 = 3 times as likely (showing the already noted brand loyalty effect).

13
Table 7 Duplication of buying at regular and promotion price in a year: Averages across Pizza
brands
Row
Buyer of
Average
brand at:
Who also buys the
column brand when
% Average over all brands
Goodfellas
Ch’go
Town
Dr
Oetker
Pizza
Express
M&S
Average
1
Regular
Diff brand at Regular
30
22
9
8
6
15
2
Promo
Diff brand at Regular
28
22
7
6
4
13
3
Regular
Diff brand at Promo
37
28
11
3
3
16
4
Promo
Diff brand at Promo
43
34
13
3
2
19
5
Regular
Same brand at Promo
65
67
51
26
22
46
6
Promo
Same brand at Regular
45
46
33
48
45
43
7
Penetration
Regular
20
16
5
4
3
10
8
Penetration
Promo
28
24
8
2
2
13
Underlying data from Kantar Worldpanel
Table 8 extends the analysis in Table 7 to all the categories in our dataset.
3.2.1 Buying different brands
Across all the categories, we see in Table 8 that on average, promotion buyers are as likely to
also buy a different brand at regular price as regular-price buyers are (rows 1 and 2 – both
19%). However, as per rows 3 and 4, promotion buyers are a bit more likely to buy a different
brand on promotion compared to regular price buyers (20 vs 17, row 4 compared to 3). This
reflects the smallish incremental effect of deal-proneness over and above the amount of
promotion available in the category. However, we can clearly also take out from this
information that buying a brand on promotion does not deter consumers from buying other
brands at normal price. The results underscore that brands are still a very important purchase
driver, consumers don’t just buy based on what is on promotion. It is also possible that some
of this small effect is due to high levels of promotion coinciding with seasonality. This would
mean less frequent buyers enter the category at its high point, and also encounter more
promotions at the time.
3.2.2 Buying the same brand
Next, in Table 8 we see figures of 40 and 41 for the average proportions buying the same brand
at Promo or Regular. Firstly, both these figures are much higher than those for buying
different brands (19,19,17,20 for rows 1,2,3,4) – this indicates the likelihood of buying the
same brand on/off promotion is, on average, around twice as high as buying other brands
on/off promotion. Second, the average figure of 40 for Regular price buyers also buying the
same brand at Promo is essentially the same as the 41 for Promo also buying the same brand at
Regular. This strongly implies that buying a brand on Promotion does not tend to diminish
buyer’s propensity to also buy it at Regular, any more than buying a brand at Regular price
would diminish their propensity to also buy it on Promotion.
As a final note to the analysis, we build on the main finding (section 3.1) that the overall extent
of promotion sales volume in a category drives the extent to which consumers only buy on
promotion. To do so, we correlate the overall promotion extent in each category (Table 8 row
7) with the proportion of regular-price buyers who buy a different brand on promotion (row 3),
and with the proportion of regular-price buyers who buy the same brand on promotion (row 5).
The correlation for different brand is .66, whereas the correlation for same brand is .84. This
finding suggests that promotion prevalence tends to result in buyers buying their same brands

14
on promotion, moreso than driving them to seek out different brands on promotion. The
general impression from these findings is that the extent of consumers actively seeking deals is
low, which is consistent with past research (Pechtl, 2004).

15
Table 8 Duplication of buying at regular and promotion price in a year: average results for 2007 & 2014 data
row
Buyer of
brand at …
Avg. % who also bought
…
Fabric
Care '14
Pizza
Inst.
Coffee
Dental
Care
Ice
Cream
Beer
Toilet
Tissue
Yoghurt
Drinks
Skin
Care
Biscuits
From.
Frais
Fabric
Care '07
Nappies
Baked
Beans
Air
Fresh
Razor
Blades
Condi
ments
Avg.
1
Regular
Diff brand at Regular
14
15
15
21
9
17
20
18
17
22
23
22
25
17
28
12
24
19
2
Promo
Diff brand at Regular
12
13
13
19
7
17
17
21
17
21
23
22
30
16
30
11
27
19
21
3
Regular
Diff brand at Promo
21
16
22
20
22
18
29
15
14
24
25
17
19
15
10
3
3
17
4
Promo
Diff brand at Promo
27
19
26
24
24
21
36
18
17
25
28
20
25
18
13
4
2
20
5
Regular
Same brand at Promo
58
46
51
48
61
43
52
43
29
45
49
47
30
34
21
9
5
40
6
Promo
Same brand at Regular
33
43
32
42
21
38
31
62
34
45
45
61
37
48
57
37
52
41
7
% sold on Promotion
53
51
51
47
45
43
35
35
35
32
29
29
28
17
16
9
7
Underlying data from Kantar Worldpanel

4. Concluding Discussion on Brand Loyalty and Deal Proneness
Brand loyalty, in terms of biasing purchases to a small personal repertoire and therefore
returning to the same brands over and over, has been documented in a wide range of product
and service categories (Ehrenberg et al., 2004), durables (Bennett, 2008), luxury goods
(Romaniuk and Sharp, 2016), and in business-to-business buying (Pickford and Goodhardt,
2000). The patterns of brand loyalty are therefore quite law-like (Ehrenberg, 1993). Such
loyalty also been observed beyond buying, for example in TV program choice (Sharp et al.,
2009) and even university lecture theatre seating (Livaditis et al., 2012). It has been strongly
suggested that it is natural human behavior, rather than marketing induced (Sharp et al., 2013).
Our findings provide further support that loyalty is a fundamental aspect of buying and perhaps
all human choice behavior. In the face of regular attractive price discounts, brand loyalty is
still very starkly apparent.
It is commonly, and not unreasonably, thought that buying a brand on deal is the antithesis of
brand loyalty. However, our data shows that even in categories where more than half of sales
are on deal there is a high degree of brand loyalty. Indeed our results indicate price promotions
gain a large proportion of their sales from people who already include the brand within their
repertoire. This finding closely corroborates past findings by Ehrenberg, Hammond and
Goodhardt (1994) that showed the majority of people who buy a brand on price promotion had
already bought the brand in the recent past.
This study has uncovered three main patterns in buying brands on and off promotion, which
have not been previously investigated. First, that the tendency for a high proportion of brand
buyers to buy a brand only when it is on promotion is directly related with the total amount of
sales bought on promotion for a brand. Next, those who buy a brand on promotion are a little
more likely than regular-price buyers to buy other brands when they are on promotion, but
buying a brand on promotion does not lower a shopper’s likelihood of buying other brands at
normal price. Third, more volume sold on deal has a stronger relationship with regular-price
buyers buying the same brand on promotion, compared to buying different brands on
promotion.
Price promotion therefore does not create big divisions (segments) in markets between the
deal-prone and those buyers who are not, but rather creates the possibility for most buyers to
take advantage of promotions when they are available. Buying on deal is a reflection of the
tendency of brands to sell a lot on promotion rather than of consumers to seek out deals. Here
we have examined this issue across 16 diverse CPG categories, but extending the analysis to
even more categories and countries would strengthen the conclusion.
5. Limitations & directions for future work
Our study examined behaviours, not price promotions. We documented mainly cross-sectional
differences in buying between brands and categories with differing levels of promotion and
normal sales. This somewhat limits our ability to comment on whether increasing amounts of
price promotion encourage sensitivity to price promotion (e.g. learning effects). This is
certainly a direction for further work.

17
In this study we undertook one comparison over time for a category (Fabric Care) for which
we had data for two time periods, seven years apart. It would be fascinating and enlightening
to carry out a full longitudinal study to examine how the incidence of households buying
categories and brands only on deal, only on promotion, or both might change in cases where
brands have changed their price promotion strategy. However, finding cases of significant
changes in price promotion strategy, particularly reductions in promotion incidence, may prove
difficult.
Lastly, our measure of promotion prevalence was the total proportion sold on deal (promotion)
in each category. An alternative measure of promotion prevalence is the amount of promotion
offered by brands over a period of a year. However, calculating such a measure would be
challenging because brands cannot be neatly classified as on or off promotion in any given
week. Rather, any brand is likely to have some of its SKUs or variants on promotion, in some
retailers and not others, in any given week. It is therefore not possible to simply add up the
proportion of weeks brands are on or off promotion. With that in mind, a direction for future
research is to develop a way to calculate this prevalence or availability of price promotions for
a category, taking into account multiple SKUs per brand and some being on deal or not at any
given time, differing levels of brand market shares, and multiple retailers. This measure could
then be used in future examinations of the interplay between promotion availability, total sales
sold on deal, and the extent of households buying on and off promotion.

18
References
Bell, David R. & Lattin, James M (2000). Looking for Loss Aversion in Scanner Panel Data:
The Confounding Effect of Price Response Heterogeneity, Marketing Science, 19, 2,
185-200
Bennett, Dag (2008). Brand loyalty dynamics - China's television brands come of age,
Australasian Marketing Journal, 16, 2, 39-50
Blattberg, R C, Buesing, T, Peacock, P & Sen, S K (1978). Identifying the deal prone segment,
Journal of Marketing Research, 15, 3, 369-377
Bogomolova, Svetlana, Dunn, Steven, Trinh, Giang, Taylor, Jennifer & Volpe, Richard J.
(2015). Price promotion landscape in the US and UK: Depicting retail practice to
inform future research agenda, Journal of Retailing and Consumer Services 25, 1, 1-11
Bolton, Ruth, Shankar, Venkatesh & Montoya, Detra (2010). Recent trends and emerging
practices in retailer pricing KRAFFT, M. & MANTRALA, M. K. (eds.), Retailing in the
21st Century: Current and Future Trends., Springer-Verlag, Berlin, 245-359.
Dawes, John (2004). Assessing the impact of a very successful price promotion on brand,
category and competitor sales, Journal of Product & Brand Management, 13, 5, 303-
314
Dawes, John G. (2016). Testing the robustness of brand partitions identified from purchase
duplication analysis, Journal of Marketing Management, 32, 7, 695-715
Dawes, John, Meyer-Waarden, Lars & Driesener, Carl (2015). Has brand loyalty declined? A
longitudinal analysis of repeat purchase behavior in the UK and the USA, Journal of
Business Research, 68, 2, 425-432
Dawes, John & Nenycz-Thiel (2014). Comparing retailer purchase patterns and brand metrics
for in-store and online grocery purchasing, Journal of Marketing Management, 30, 3-
4, 364-382
Dekimpe, Marnik G., Hanssens, Dominique M. & Silva-Risso, Jorge M. (1998). Long-run
effects of price promotions in scanner markets, Journal of Econometrics, 89, 1-2, 269-
291
Dodson, Joe A., Tybout, Alice M. & Sternthal, Brian (1978). Impact of deals and deal retraction
on brand switching, Journal of Marketing Research, 15, 1, 72-81
Drechsler, Salome, Leeflang, Peter, Bijmolt, Tammo Hendrik Antonie & Natter, Martin (2017).
Multi-unit price promotions and their impact on purchase decisions and sales,
European Journal of Marketing, 51, 5/6, 1049-1074
Ehrenberg, Andrew (1993). Even the social sciences have laws, Nature, 365, 30, 385
Ehrenberg, Andrew (2000). Repeat-buying: facts, theory and applications, Journal of
Empirical Generalisations in Marketing Science, 5, 2, 392-770
Ehrenberg, Andrew & Goodhardt, Gerald (1968). Incidence of brand-switching, Nature, 220,
304
Ehrenberg, Andrew & Goodhardt, Gerald (1969). Practical applications of the duplication of
viewing law, Journal of the Market Research Society, 11, 1, 6-24
Ehrenberg, Andrew, Hammond, Kathy & Goodhardt, Gerald J (1994). The After-Effects of
price-related consumer promotions, Journal of Advertising Research, 34, 4, 11-21

19
Ehrenberg, Andrew, Uncles, Mark D & Goodhardt, Gerald G (2004). Understanding brand
performance measures: using dirichlet benchmarks, Journal of Business Research, 57,
12, 1307-1325
Goodhardt, Gerald & Ehrenberg, Andrew (1969). Duplication of viewing between and within
channels, Journal of Marketing Research, 6, 2, 169-178
Hardie, Bruce G.S., Johnson, Eric J. & Fader, Peter S. (1993). Modeling Loss Aversion and
Reference Dependence Effects on Brand Choice, Marketing Science, 12, No. 4, Fall,
378-394
Johnson, Tod (1984). The myth of declining brand loyalty, Journal of Advertising Research,
24, 1, 9-17
Jones, John Philip, (1990). The Double Jeopardy of Sales Promotions, Havard Business
Review.
Kalyanaram, Gurumurthy & Winer, Russell S. (1995). Empirical generalizations from
reference price research, Marketing Science, 14, 3, 161-169
Kantar, (2015), Kantar Worldpanel: Consumer panels [Online],
http://www.kantarworldpanel.com/global/Consumer-Panels,
Klapper, Daniel, Ebling, Christine & Temme, Jarg (2005). Another look at loss aversion in
brand choice data: Can we characterize the loss averse consumer, International
Journal of Research in Marketing, 22, 239-254
Lichtenstein, Donald R., Netemeyer, Richard G. & Burton, Scot (1995). Assessing the domain
specificity of deal proneness: a field study, Journal of Consumer Research, 22, 3, 314-
326
Livaditis, M., Sharp, Anne & Sharp, Byron. (Year). Evidence of Naturally Bias Behaviour -
Seating Habits at a Lecture. ANZMAC, 2012, Adelaide, Australian and New Zealand
Marketing Academy.
Martînez, Eva & Montaner, Teresa (2006). The effect of consumer's psychographic variables
upon deal-proneness, Journal of Retailing and Consumer Services, 13, 157-168
Mazumdar, Tridib, Raj, S. P. & Sinha, Indrajit (2005). Reference price research: Review and
propositions, Journal of Marketing, 69, October, 84-102
Mela, Carl F., Gupta, Sunil & Lehmann, Donald R. (1997). The Long-Term Impact of
Promotion and Advertising on Consumer Brand Choice, Journal of Marketing
Research, 34, May, 248-261
Mela, Carl F., Jedidi, Kamel & Bowman, Douglas (1998). The Long-Term Impact of
Promotions on Consumer Stockpiling Behavior, Journal of Marketing Research, 35,
May, 250-262
Nielsen, (2009), Almost half of U.S. supermarket purchases are sold on promotion [Online],
New York, NY, Nielsen, nielsen.com/us/en/insights/news/2009/almost-half-of-u-s-
supermarket-purchases-are-sold-on-promotion.html, [Accessed August 2016
Nijs, Vincent R., Dekimpe, Marnik G, Steenkamp, Jan-Benedict E.M. & Hanssens, Dominique
M (2001). The category-demand effects of price promotions, Marketing Science, 20, 1,
1 - 22
Palazon, Mariola & Delgado-Ballester, Elena (2011). The expected benefit as determinant of
deal-prone consumers' response to sales promotions, Journal of Retailing and
Consumer Services, 18, 6, 542-547
Pechtl, Hans (2004). Profiling intrinsic deal proneness for HILO and EDLP price promotion
strategies, Journal of Retailing and Consumer Services, 11, 4, 223-233

20
Pickford, Chris & Goodhardt, Gerald. (Year). An empirical study of buying behaviour in an
industrial market. AM2000, Academy of Marketing Annual Conference, 5-7 July,
2000, University of Derby, Derby, University of Derby, pdf copy on server.
Romaniuk, Jenni & Dawes, John (2005). Loyalty to price tiers in purchases of bottled wine,
Journal of Product & Brand Management, 14, 1, 57-64
Romaniuk, Jenni & Sharp, Byron (2016). And finally, a bit of luxury ROMANIUK, J. &
SHARP, B. (eds.), How Brands Grow: Part 2, Oxford University Press, Melbourne,
203-221.
Rossiter, John R. & Bellman, Steven (2005). Marketing communications: Theory and
applications, Pearson Education, Frenchs Forest.
Sharp, Byron (2010). How brands grow SHARP, B. (ed.) How Brands Grow, Oxford
University Press, Melbourne, Australia, 16-27.
Sharp, Byron (2013). Marketing: Theory, evidence, practice, Oxford University Press,
Melbourne.
Sharp, Byron, Beal, Virginia & Collins, Martin (2009). Television: Back to the future, Journal
of Advertising Research, 49, 2, 211-219
Sharp, Byron, Bogomolova, Svetlana, Livaditis, Marianthi & Nenycz-Thiel, Magda (2013).
Consumer behaviour and business buyer behaviour SHARP, B. (ed.) Marketing:
Theory, evidence, practice, Oxford University Press, Melbourne, Australia,
Sharp, Byron, Wright, Malcolm, Dawes, John, Driesener, Carl, Meyer-Waarden, Lars, Stocchi,
Lara & Stern, Philip (2012). It’s a Dirichlet world: Modeling individuals’ loyalties
reveals how brands compete, grow, and decline, Journal of Advertising Research, 52,
2, 203-213
Tanusondjaja, Arry, Nenycz-Thiel, Magda & Kennedy, Rachel (2016). Understanding shopper
transaction data: how to identify cross- category purchasing patterns using the
duplication coefficient, International Journal of Market Research 58, 3, 1-12
Totten, John C & Block, Martin (1994). Analyzing Sales Promotions, Dartnell Corporation,
Chicago.
Webster, Frederick E. (1965). The "Deal-Prone" Consumer, Journal of Marketing Research, 2,
May, 186-189