Journal of Transport and Supply Chain Management

ISSN: (Online) 1995-5235, (Print) 2310-8789

Page 1 of 11 Original Research

Investigating stock-outs in Johannesburg’s

warehouse retail liquor sector

Authors: Purpose: The aim of this exploratory study was to investigate how management practices and
Amit Govind1
processes contribute to stock-outs in the warehouse retail liquor sector in Johannesburg.
Rose Luke1
Noleen Pisa1 Research design: A pragmatic research philosophy was used in a multi-case approach to
Affiliations:
determine the association between business management practices and stock-outs. The unit of
1
Department of Transport analysis of the study is the highest turnover outlets.
and Supply Chain
Management, University of Findings: The results revealed that retailers base demand forecasting on judgement and naïve or
Johannesburg, South Africa simple moving average forecast methods. There is little consideration of variability, lead time or
targeted customer service levels when determining optimal inventory levels. Product breakages
Corresponding author:
are common, implying that data inaccuracies and stock-outs are highly probable. Information
Rose Luke,
[email protected] sharing between the retailers and suppliers is limited and formal collaboration programmes do
not exist. The respondents revealed that frequent stock-outs resulted in poor business
Dates: performance characterised by declining sales, customer retention and competitive advantage.
Received: 23 Mar. 2017
Accepted: 13 July 2017 Limitations: Although the sample selected represented a significant portion of the top 50%
Published: 21 Sept. 2017 contributors to the sector’s sales, the sample is small and the focus of the study is limited,
How to cite this article: which decreases the generalisability of the results.
Govind, A., Luke, R. & Pisa,
Practical implications: Retailers can reduce stock-outs by investing in improvements in
N., 2017, ‘Investigating
stock-outs in Johannesburg’s demand forecasting, synchronisation through collaboration and training of employees.
warehouse retail liquor
sector’, Journal of Transport Originality or value: This study provides empirical evidence of the linkages between poor
and Supply Chain management practices and processes, stock-outs, customer service and business performance.
Management 11(0), a303.
https://doi.org/10.4102/
jtscm.v11i0.303
Introduction
Copyright: Stock-out situations are scenarios where stock or final inventory in a retail outlet is not available
© 2017. The Authors.
Licensee: AOSIS. This work
on the shelf for a customer to purchase (Aastrup & Kotzab 2009; Grant & Fernie 2008). Customers
is licensed under the will do one of the following when faced with a stock-out in a retail store: (1) switch stores to locate
Creative Commons the same product, (2) change brands in order to get a similar utility out of the same product, (3)
Attribution License.
postpone the purchase to a timing when the product becomes available or (4) completely drop the
purchase, resulting in a lost sale to the store (Van Woensel et al. 2007). In a survey of customers in
three Dutch grocery stores, about 84% of customers in the perishable products sector were found
to substitute their product purchase when faced with a stock-out situation owing to the immediacy
effect (Van Woensel et al. 2007). The immediacy effect is defined as customer demand for a product
for immediate consumption.

Customers may switch to another brand in the short term for a product that is highly substitutable
in the same store; thus, the outlet benefits by not losing a sale (Sloot, Verhoef & Franses 2002; Van
Woensel et al. 2007). If continuous stock-outs occur, consumers will switch stores permanently,
leading to a diminishing customer base, poor customer retention, loss of store loyalty, reduction
in sales turnover and diminishing profits in the long run (Van Woensel et al. 2007). Customer
retention is a basic building block for a successful retail business in today’s competitive landscape
(McNaghten & Passino 2011). For a retail business to operate sustainably, they cannot afford
losing customers to competitors.

Stock-outs alter customer behaviour and buying patterns and increase demand variability (Van
Read online: Woensel et al. 2007). Customer behaviour and buying patterns are critical predictors and
Scan this QR
code with your influencers of sales turnover (Pearson 2011). Stock-outs make the monitoring of customer demand
smart phone or
mobile device Note: This article is partially based on the authors’ article presented and published as part of the Proceedings of the 21st International
to read online. Symposium on Logistics (ISL 2016), available here: http://www.isl21.org/wp-content/uploads/2016/09/Final-proceeding-with-abstract-
update-OCT-05.pdf

http://www.jtscm.co.za Open Access

 Page 2 of 11 Original Research

patterns difficult as brand switching takes place, which questions, this article makes two contributions. This study
causes increased customer demand variability (Corsten & extends the existing literature by synthesising earlier research
Gruen 2003). Customer demand variability is defined as a areas, namely supply chain management practices and stock-
deviation from a supply chain process (Coyle et al. 2012). outs, to illustrate the effects of supply chain management
practices on stock-outs. In addition, this exploratory research
Business management levers such as key strategic management offers empirical evidence of such a phenomenon from the
processes in the supply chain, store operations, category warehouse retail liquor businesses in Johannesburg. The
management and marketing can improve retailers’ competitive remainder of this article is organised as follows: the ‘Literature
positions and be sources of sustainable growth (PWC 2012). review’ section provides the theoretical background of
These levers are the building blocks for a competitive advantage the study; in particular, it describes the beer game and the
in the retail industry because they influence value through bullwhip effect. The ‘Research method’ section describes the
function and cost (Jones & George 2009; PWC 2012). The beer research methods employed in the study. The ‘Results and
game categorises three management levers: the management discussion’ section discusses the findings of this exploratory
of demand variability, supply chain synchronisation and study, while the ‘Conclusions and recommendations’ section
collaboration (communication). Inefficiencies in these levers concludes the article.
have been found to be probable causes of stock-outs in retail
business (Van Woensel et al. 2007).
Literature review
The retail industry in South Africa has shown a slow-down of Supply chain synchronisation is the integration of
sales growth over the span of a year from 3.9% in 2012 to 2.8% operations to achieve a mutual goal between supply chain
in 2013 because of household consumption spending that had entities (Simatupang, Sandroto & Lubis 2004). A lack of
been inhibited (Maswanganyi 2013; Reuters 2013). The synchronisation of this process flow results in variability in
reduction in household consumption spending in South the supply chain (Corsten & Gruen 2003; Sahay & Mohan
Africa was mainly attributed to inflationary pressures arising 2003; Van Woensel et al. 2007). Variability is a deviation
from higher electricity and fuel prices (Chisadza et al. 2013). from the integrated activities of the supply chain process
Many retail businesses in South Africa are facing the flow (Anupindi et al. 2012; Croxton et al. 2002). Variability
phenomena of stock-outs which are resulting in enormous occurs when there is non-synchronisation of process flows
amounts of lost sales because of customers’ demand not being mainly because of lack of communication and poor
matched with supply (Pearson 2011). According to Pearson inventory management decisions (Corsten & Gruen 2003;
(2011), the problem of stock-outs for all South African retail Sahay & Mohan 2003; Van Woensel et al. 2007). High levels
businesses has been consistent over the past 10 years and has of variability experienced in process flows for products,
not improved. Furthermore, Pearson (2011) highlighted that information and finances are correlated to either stock-outs
the problem of retail stock-outs can amount to a loss of 4% of or over supply in the retail industry (Aastrup & Kotzab
the annual sales turnover for an average retail business in 2009; Corsten & Gruen 2003; Miranda & Jegasothy 2008;
South Africa. This lost sales turnover in the retail sector has Pramatari & Miliotis 2008). Variability in a supply chain
several short-term and long-term consequences for a process causes uncertainty and creates unpredictable
business’s profitability and survival. In the liquor retail process flows and non-synchronisation of product flows,
sector, the problems are exacerbated by inflationary information exchange and finances in the supply chain
pressures, consolidation of outlets (creating barriers to entry (Anupindi et al. 2012).
to the market and market positioning problems) and the high
level of unlicensed liquor outlets, because of complex licensing A collaborative arrangement between suppliers and retailers
requirements (PWC 2012). These pressures, especially for is necessary to attain synchronisation and agility in the
unlicensed outlets, make forecasting in the sector particularly supply chain. ‘Agility’ refers to suppliers’ ability or flexibility
difficult – unlicensed outlets are often raided by police, to react to retailers’ needs. This is facilitated by the sharing of
reducing demand in an area for a while, only to resume a few stock or sales information between the retailer and the
months later, thereby impacting demand forecasting. As there suppliers through formal collaboration programmes (Sahay
are approximately 120 000 unlicensed liquor outlets in the 2003). Collaboration with suppliers is crucial to creating the
country, this can significantly impact demand. Poor demand synergies that are required to reduce variability in the supply
forecasting and stock-outs in the sector are confirmed by chain (Pawlak & Malyszek 2008; Sahay 2003). The lack of
observation by one of the researchers, in his position as strategic relationships and collaboration with suppliers leads
demand planner for the largest beer manufacturer in Africa. to increased variability and stock-outs at the retail outlets.
Stock-out situations have a negative effect on working capital
By extension, it is conceivable that new insights might be for a business (Wisner, Tan & Leong 2012). Stock-outs and the
gained by considering that an association exists between associated low inventory levels result in lost sales as products
business management levers in the form of supply chain are not available when the customer wants to make a
management practices and processes and stock-outs. The aim purchase. Continued stock-out situations lead to customer
of this exploratory study is to investigate business management dissatisfaction, loss of customer loyalty and, consequently,
practices that contribute to stock-outs in warehouse retail a decline in profit margins, and competitive advantage
liquor businesses in Johannesburg. In addressing these (Atrill & McLaney 2008).

http://www.jtscm.co.za Open Access

 Page 3 of 11 Original Research

To explore the dynamic nature of stock-outs in the supply beer game, supply chain entities are exposed to batch ordering
chain, two broad streams of research are particularly and adjust ordering patterns in an attempt to match demand
informative: the beer game and the bullwhip effect. with supply. These attempts to adjust ordering patterns are
The former explores the extent to which variability, known as reactive decisions in the form of operating policy
synchronisation and collaboration characterise business formulation (Goodwin & Franklin 1994). The existence of
management processes and practices and the impact of these operating policies illustrates managerial inefficiencies in
factors on business performance. The latter provides the customer demand signalling, synchronisation, collaboration
basis for understanding demand variability. Combined, these and inventory stocking policies (Lee et al. 1997). The operating
two streams of research provide an understanding of the policies (that participants of the supply chain process flow
association between business management processes and establish in the beer game) result in the bullwhip effect which
practices and stock-outs. increases variability in the supply chain. Ultimately, the beer
game illustrates the effect of poor synchronisation and
communication in a supply chain process.
The beer game
A retail organisation can be faced with stock-outs if there is a The beer game highlights that demand forecasting, safety
lack of synchronisation of customer demand signals from stock modelling and inventory replenishment policies are
one supply chain entity to the next in the process flow critical inventory management processes and practices that
(Goodwin & Franklin 1994). The lack of communication of influence synchronisation and the management of variability
demand signals can be explained using the beer game in an organisation (Goodwin & Franklin 1994). Poor demand
simulation. The concept of the beer game was originally forecast methods from the retailer cause a bullwhip effect
developed in the 1960s by Massachusetts Institute of through the supply chain. In addition, poor communication
Technology’s Sloan School of Management students, as an among supply chain entities leads to poor inventory
experiment to illustrate supply chain communication and buffering, which results in stock-outs. Batch ordering is a
lack of synchronisation of demand signals. The beer game result of poor inventory replenishment ordering. Effective
simulation illustrates the process of entities of a supply chain inventory management is dependent on the application of
receiving and processing information and products. It effective demand forecasting techniques (Croxton et al. 2002;
comprises four participants with the following positions: Goodwin & Franklin 1994). Customer service levels can be
manufacturer, distributor, wholesaler and retailer. Two rules ensured by buffering inventory through safety stock
are applied in the game. Firstly, there should be no modelling and applying the economic order quantity
collaboration or communication between positions when principle of inventory management (Jeffery, Butler & Malone
orders or demand signals are transmitted from downstream 2008). These methodologies achieve optimisation of inventory
to upstream. Secondly, there is a two-period lead time for management to fulfil customer demand and reduce the
ordering and signalling between each supply chain entity. probability of a stock-out (Croxton et al. 2002; Goodwin &
Franklin 1994; Jeffery et al. 2008).
To illustrate the concept, a customer places an order of 100
cases with the retailer, but the retailer has only 80 cases in
stock. This implies that there is a shortage in supply of 20
Bullwhip effect
cases. The retailer will only receive the additional 20 cases of Supply chain variability can also be characterised as the
stock from its wholesaler in two periods’ time. Therefore, in bullwhip effect (Lee et al. 1997). This was originally
the current period, the retailer has 80 cases in stock and a conceptualised by Forrester (1961) who showed that ‘small
stock-out of 20 cases. The lack of synchronisation between changes in retail sales can lead to larger swings in factory
retailer and wholesaler can be simulated further upstream in production’ owing to the internal characteristics of the
the supply chain process. A similar occurrence could take system. The bullwhip effect is defined as amplified customer
place with the wholesaler and the distributor, thus illustrating order variability in the supply chain, as orders move up the
the concept of poor matching of demand for goods with the supply chain. Increased orders through the supply chain are
supply of goods (Croxton et al. 2002; Lee, Padmanabhan & caused by each entity ordering for their own gain. Each entity
Whang 1997). in the supply chain may increase the order volume to earn
quantity discounts, thus distorting the initial customer order
This poor matching of demand and supply leads to back information (Lee et al. 1997; S. Parsonsons, pers. comm.,
ordering, in order to fulfil the original demand deficit of 20 25 September 2013). The impact of the bullwhip effect is the
cases. According to Lee et al. (1997), this back ordering is transfer of distorted information to each supply chain entity
termed demand forecasting updating or demand signal resulting in decreased efficiencies. This distorted information
processing, where the retailer experiences a surge in demand exchange results in excessive inventory or stock-outs and
and signals to the upstream supply chain entity to increase subsequent poor customer service.
the next order to be larger than normal. This larger than
normal size is termed batch ordering (Lee et al. 1997). Batch Variability occurs when incorrect or inaccurate customer
ordering is triggered to cover future demand signals from demand data are communicated throughout the supply
customers that cannot be fulfilled with the retailer’s current chain, resulting in incorrect management decisions (Bon &
available stock (Rungtusanatham et al. 2007). Throughout the Leng 2009). A lack of customer demand data occurs when

http://www.jtscm.co.za Open Access

 Page 4 of 11 Original Research

supply chain entities in the process flow do not receive Research design
information in time or do not receive it at all for critical
A pragmatic research philosophy was used in a multi-case
business decision-making processes (Lee et al. 1997;
approach to determine the association between business
Pawlak & Malyszek 2008). Therefore, understanding the
management processes and practices and stock-outs. The
effects this has on entities in the supply chain is crucial in
pragmatic approach sanctions the utilisation of both
managing business processes which could enhance
quantitative and qualitative methods (Onwuegbuzie & Leech
process flow synchronisation (Fliedner 2003; Ghosh &
2005; Onwuegbuzie & Teddlie 2003). Therefore, a concurrent
Fedorowicz 2008).
mixed methods approach was used to collect both quantitative
data and qualitative data. Quantitative data on measurable
In categorising customer demand variability, the bullwhip
variables, for example, the frequency of stocks, were collected.
effect is caused by the following factors: demand forecast
The inclusion of quantitative data also helps to compensate
updating, order batching, price fluctuations, rationing and
for some of the shortcomings of qualitative data, namely that
shortage gaming (Kalchschmidt, Verganti & Zotteri 2006; it cannot be generalised (Onwuegbuzie & Leech 2005). In
Lee et al. 1997). According to Lee et al. (1997), shortage addition, qualitative data on the non-measurable variables
gaming is the act of a customer anticipating the business to were collected using open-ended questions in order to obtain
be out of stock which causes an earlier purchase to gain a deeper understanding of the phenomenon and to answer
stock. Coyle et al. (2012), Kalchschmidt et al. (2006) and Lee the research questions from the cases. Case-based approaches
et al. (1997) found that customer demand is clustered have been shown to contribute to the advancement of theory
according to types of products and patterns of customer by providing in-depth insights and understanding of
demand purchases. Lee et al. have described variability in complex areas of management practice (Yin 2014). In
the supply chain as a lack of synchronisation driven addition, interviews were used to obtain rich descriptions
from customer demand signals. Hence, understanding from the alcohol retailers. In order to improve the
these products and customers is essential to creating generalisability of our findings, five general methodological
synchronisation in the supply chain. phases were employed. Firstly, three research questions were
formulated; secondly, a questionnaire was developed; thirdly,
Research method field data were gathered; fourthly, interviews were
transcribed and analysis performed using coding schemes
Case selection and cross-case comparisons; and, finally, findings were
Johannesburg was selected because it is the main economic disseminated to improve validity.
hub of South Africa. The focus of this study was formal
trade warehouse retail outlets that sell to other retailers as
Data sources and collection
well as end-users (consumers). Formal trade is a category
of the alcohol beverage industry in South Africa, where Data were collected during interviews using semi-structured
these businesses operate by keeping stock in their questionnaires. The responses were scaled on Likert and
warehouses and trade to formal and informal customers category scaling systems in order to drive a clear answer
daily (Herrick & Parnell 2013). The population of the when questions are rated (Zikmund 2003). The interviews
warehouse retail liquor outlets in Johannesburg comprises were conducted with managers in the areas of procurement,
approximately 917 outlets, which is 14.8% of the total sales and supply chain management from the different firms
outlets selling alcoholic beverages in Johannesburg. in the sample. The data were collected over an 8-month period
between March and October. This is assumed to be a
Because of the large size of the population (917 outlets), it
reasonable period of time for an indication of stock-out causes.
was stratified according to sales volume turnover per
The assumption is based on a management interview with an
month. The highest turnover strata were selected for
alcohol beverage manufacturer in South Africa who revealed
further analysis because these strata would typically be
that the 8 months are indicative of typical sales patterns of
more prone to stock-outs. This is because of the higher
retail outlets (C. Kubeka, pers. comm., 02 July 2013).
frequency of store ordering and replenishment required to
fulfil the high-volume turnover (Corsten & Gruen 2003).
Fifty per cent of the 917 outlets’ volume turnover was Results and discussion
derived from approximately 70 outlets, indicating that The total sample size interviewed for the case study was 25
there were a large number of outlets with low volume retailers; however, 22 responses were considered valid for the
turnover. The target population of 70 high sales volumes analysis. The other three questionnaires were discarded from
outlets was further stratified according to the extent of the analysis as these were incomplete or invalid. Management
homogeneity of the population group. The homogeneity of processes and practices were analysed to determine whether
the population group was based on a similar trading and inefficiencies were present and whether those inefficiencies
operating format to customers in the formal trade resulted in stock-outs in the retail outlets. In the beer game,
(R. Alborough, pers. comm., 20 September 2013; C. Kubeka, critical outcomes of the simulation were revealed about a
pers. comm., 02 July 2013; S. Parsonsons, pers. comm., supply chain that relates back to management processes and
25 September 2013). Resultantly, 35% or 25 outlets of the practices which then impact stock-outs at retail businesses.
high-volume outlets were targeted. Based on the conceptual framework and the literature review,

http://www.jtscm.co.za Open Access

 Page 5 of 11 Original Research

findings were separated into three areas: variability, 20 years of experience, and in all these cases respondents
synchronisation and communication. indicated that the business was family owned and operated.
This is consistent with the literature that indicates that
forecasting customer demand using the judgemental
Variability
approach requires prior knowledge. The use of simple
The beer game illustrates that poor customer demand forecasting methods implies a high probability of sending
signalling can create the bullwhip effect through the supply poor customer demand signals to suppliers when placing an
chain. The bullwhip effect results in increased variability, inventory order (Coyle et al. 2012; Wisner et al. 2012). As a
from retailer level through to supplier, as incorrect information result, stock-outs would be highly probable. This is consistent
is passed through the supply chain by every entity. Demand with the beer game, as poor order signals through the supply
forecasting methodologies, customer demand patterns chain results in poor matching of demand with supply.
tracking or forecast tracking and inventory management
have been shown to influence demand variability and to be Respondents indicated that trends and seasonality had not
primary reasons for stock-outs. These were investigated in factored into demand forecasting techniques. Hence, this
this study and the following results were deduced. indicates that failing to implement more sophisticated
forecasting techniques during seasonal customer demand
Forecasting methodologies patterns may result in stock-outs.
In order to determine the forecasting methodologies
employed by the retail outlets, the respondents were asked to Eighty-three per cent of respondents rated ‘no’ to ‘low’
identify the methods used in their business. Each of the adherence to cause and effect modelling as a forecasting
possible forecasting methods indicates the level of technique. This implies that customer demand events are a
sophistication. The key finding was that customer demand probable cause of stock-outs. Events such as changing
forecasting was mostly based on judgement, naïve or simple weather patterns, sporting events, music festivals and long
moving average forecast methods (see Figure 1). These weekends were cited as customer demand events that cause
forecasting methods are limited and less reliable in forecasting stock-outs because of difficulty in predicting customer
customer demand in different situations. There was a general demand. Cause and effect modelling is a supporting tool for
lack of use of sophisticated demand forecasting techniques. a business that can determine a relationship with customer
Of the respondents, 50% did not use statistical software to demand through regression analysis (Frank et al. 2003;
forecast customer demand. Only 9% of the respondents Wisner et al. 2012). This indicates that failing to implement
recorded high adherence to the use of statistical software to modelling techniques for customer demand events and
forecast customer demand, as shown in Figure 1. considering local issues may result in stock-outs.

Of the respondents, 73% used judgement or intuition to Customer demand patterns tracking or forecast tracking
determine inventory orders. This could be attributed to the The poor forecasting methods result in a lack of understanding
number of years of experience in the business, making of customer demand for the retailer and lead to poor
informed customer demand forecasting decisions based on inventory planning. There appeared to be tracking of
prior knowledge, as suggested by Nakano and Oji (2012). customer demand data on excel spread sheets and
The study found that 75% of respondents indicated that the computerised systems (capturing electronic point of sale
business was family owned and had been operating for data). Forty-one and twenty-seven per cent of respondents
10 years or more. Of these respondents, 36% had more than indicated ‘medium’ and ‘high’ adherence, respectively, to

No
80 Low
73 Medium
70 68 High

60
50
Percentage (%)

50

40

30
23 22
20 18 18
14
10 9
0 5
0
0
Stascal sowares to forecast Cause and effect modelling Judgement or intuion of inventory
customer demand order
Forecast methods

FIGURE 1: Forecast methods used (respondents’ rating in %).

http://www.jtscm.co.za Open Access

 Page 6 of 11 Original Research

tracking of customer demand data on excel spread sheets. increased customer demand variability, according to the
However, there was low adherence to the application of data literature by Lee et al. (1997), may lead to stock-outs.
analysis or forecast tracking methodologies. By tracking
forecast errors, businesses would be able to switch to different Customer demand predictability
forecasting techniques to reduce variability. Sixty-eight per Respondents were asked to rate the level of predictability of
cent of respondents captured electronic point of sales data, customer demand patterns on a four-point scale, from 1 being
while 32% did not do so. Respondents indicated that data low predictability to 4 being high predictability. Figure 3
were used for naïve or simple moving customer demand illustrates customer demand predictability across different
forecasting when ordering products from suppliers. The data conditions is fairly low. Although customer demand tracking
contained historical customer sales and stock information data are recorded, advanced data modelling or analysis is
using last period customer sales to determine the forecast for poor, as shown in Figure 1. There appears to be an association
the next period. The store ordering procedures were based on between the lack of analysing customer demand data and the
this methodology. low level of customer demand predictability in the outlets.
The same needs of customers under the variable in the
Forty-six per cent of respondents recorded no and low questionnaire titled homogeneous customers were rated
adherence (14% and 32%, respectively) to safety stock fairly high in terms of customer demand predictability which
modelling (see Figure 2). Respondents indicated that all was expected as repeated buying patterns are easier to track
inventory in the retail outlets was determined based on a and predict. Ninety-six per cent (64% and 32%, respectively)
standard stock cover measure. There appeared to be no of respondents rated heterogeneous customers, that is,
customer demand variability factored into the safety stock different needs of customers, to have ‘no’ or ‘low’ impact on
modelling techniques implemented by respondents. The customer demand predictability. Heterogeneous customer
primary reason for this was a poor understanding of customer demand patterns were further divided into customer
demand. According to Jeffery et al. (2008), stock cover is a behaviour categories.
ratio of closing stock to customer demand, which can be
expressed in days or weeks. Respondents indicated that a Thirty-two per cent of respondents indicated that promotional
stock cover of between 1 and 2 weeks was kept as safety stock activity has no impact on predictability of customer demand,
for all product lines at all times. while 41% indicated it as having a low impact. These
occasions also resulted in stock-outs. The seasonality
Respondents indicated that they had implemented a simpler category was reviewed in the questionnaire in terms of
version of lead time, customer demand variability and seasons of the year that are prone to increased purchasing,
customer service levels based on a computerised stock the seasons being Christmas and Easter or summer and
system. This computerised stock system considers previous winter in South Africa. Sixty-eight per cent of respondents
week customer demand and current levels of stock, which rated seasonal customer demand to have low predictability
calculates stock cover ratio expressed in weeks. Respondents on customer demand. The literature indicates that it is
further indicated that 2 weeks’ stock cover based on this unlikely to accurately predict customer demand during
computerised system was their safety stock policy. The seasonal periods without using appropriate forecasting
method does not take trend and seasonality into account, methods. Likewise, it appears that poor customer demand
which may increase customer demand variability. The forecasting methodologies are implemented, which results in

60 No
Low
50 Medium
50 High

41 40
40
Percentage (%)

32 32
30 27
23

20 18
14 14

10 9

0
0
Excel sheet tracking demand Electronic point of sale data Safety stock modelling (lead me,
captured demand variability and service level)
Forecast tracking methods

FIGURE 2: Forecast tracking methods (respondents’ rating in %).

http://www.jtscm.co.za Open Access

 Page 7 of 11 Original Research

No
Low
80
Medium
High
70 68
64
60
55
Percentage (%)

50
41 41
40 36
32 32 32 32
30 27 27 27 27
23
20 18

10
4 4 5 5
0 0 0 0
0
Homogeous Heterogenous Promoonal New product Seasonality e.g. Price discounts
customers customers acvity e.g. launches xmas, Easter,
scratch and win summer
Demand predictability measures

FIGURE 3: Customer demand predictability (respondents’ rating in %).

stock-outs. This is confirmed by the chosen inventory Product item data accuracy
management techniques, which tend to be judgement or Respondents were asked to rate weekly stock count process,
intuitive methods of forecasting (see Figure 1). classification and documentation of stock received and
updating price information in their store, from 1 being low
Brand loyalty was reviewed in the questionnaire in terms of to 4 being high adherence. Figure 4 illustrates that over 75%
new products entering the market. Brand loyalty occurs of respondents rated medium to high adherence to the
when customers maintain a long-term relationship with a stock count process, updating price information and
specific brand and do not switch to other brands easily. The classification and documentation of stock received. The
entry of new products in the market adds complexity in
medium to high adherence of these processes removes the
demand variability and increases the probability of stock-
probability of data inaccuracy, which assists in reducing
outs (Salmi & Holmström 2004). Seventy-eight per cent of
stock-outs in the store.
respondents indicated new products entering into the market
had low predictability, which suggests a probable cause of
Ordering and inventory accuracy: Figure 4 illustrates weekly
stock-outs. Customer demand forecasting methodologies are
ordering or purchasing processes, with 1 being low and 4
limited to judgement or intuition which could also be a
being high adherence. Eighty-six per cent of respondents
contributing factor in understanding and analysing the
rated high adherence to the ordering process suggesting that
customer demand data.
if ordering processes are in place in the outlets in most
instances, then it reduces the probability of stock-outs.
Synchronisation This conflicts with earlier findings that indicated the lack
Synchronisation is defined as matching demand with of understanding customer demand during seasonal,
supply throughout the supply chain. In the beer game, promotional, new products into the market, and price
synchronisation was a critical activity in the supply chain discounts. This conflict suggest that if there are periods
for each entity to fulfil their function based on an initial where customer demand is difficult to predict, the usual
input, which was an order signal from a customer. The order retail store ordering process would be inadequate in those
signal was the starting point of the supply chain and times, thus resulting in stock-outs.
therefore each entity had an onus to fulfil an order on time
and in full from upstream to the downstream customer. Store and shelf replenishment: Respondents were requested
Store shelving, store ordering and forecasting are the critical to rate, on a scale from 1 to 4, daily merchandising of shelves
factors to supply chain synchronisation (Corsten & Gruen and handling of products to avoid breakages in stores. A
2003). Respondents were asked to rate on a four-point Likert score of 1 represented poor adherence, while a score of 4
scale the adherence to retail operations process modelling, represented high adherence to this process. Figure 4 illustrates
from 1 which is poor adherence to 4 which is high adherence. that over 80% of respondents rated high adherence to daily
The retail operations process modelling is based on three merchandising of shelves. The high adherence to this process
criteria cited as reasons for stock-outs in retail, as indicated plays an important role in the final purchase decision for a
in the literature. The three criteria were product item data consumer in terms of stock availability (Baird & Rosenblum
accuracy, ordering and inventory accuracy and store and 2010). Over 60% of respondents rated handling of products to
shelf replenishment. avoid breakages as low.

http://www.jtscm.co.za Open Access

 Page 8 of 11 Original Research

Communication orders placed by a retailer are filled, according to

specifications, by the supplier. Communication of orders,
In the beer game simulation, no communication is allowed
therefore, did not present a risk of stock-outs.
between each supply chain entity. By design, this results in
each entity forecasting customer demand individually. The
lack of communication leads to overstocking or stock-outs Information sharing: Respondents were asked to rate, on a
and poor synchronisation and increased variability four-point scale, the level of information sharing using
throughout the supply chain. To test whether collaborative electronic point of sale or inventory data with suppliers.
arrangement exists between retailers and suppliers, Figure 5 illustrates that 59% of respondents rated a poor- to
respondents were asked to rate their communication with low-quality standard of data sharing. Respondents also
suppliers regarding order placing interaction, information indicated that sales and inventory data were not shared with
sharing and collaboration programmes. suppliers, as there was a lack of trust and that data or
information was strictly for the owners only. The literature
Order placing interaction indicates that trust and commitment between supplier and
Respondents were asked to rate, on a scale of 1 to 4, the ease retailer are critical factors for the success of a collaborative
of placing an order with a supplier by communication with a arrangement (Ghosh & Fedorowicz 2008). Based on the
sales representative or call centre. Ninety-five per cent of responses, it is indicative that there is a lack of trust, and
respondents rated this as being of a medium- to high-quality hence collaborative arrangements are absent, which will
standard (see Figure 5). These results suggest that affect the agility of suppliers in reacting to the needs of
communication is relatively good, as cooperation is present if the retailer.

No
Low
100
Medium
90 86 High
80
73 73
70 64
Percentage (%)

60 55 54
50
40
32
30 26
22 23
20 17
14 14 14
9
10 5 5 5 5 4
0 0 0 0
0
Weekly stock Weekly ordering/ Daily Updang of price Product handling in Classificaon and
counts purchasing cycle merchandising of informaon the storeroom to documentaon of
shelves prevent breakages stock received
Retail and operaons process modelling

FIGURE 4: Retail and operations process modelling (respondents’ rating in %).

No
100 Low
Medium
90 86
High
80
72
70 68
Percentage (%)

60
50
50 45
40 36
32 32 32
30 23 23 23 23
20
9 9 9 9 9
10 5 5
0 0 0 0
0
Ease of placing an Adherence to order Availability of Correct stock Collabora
on/ Sharing (EPOS) data
order with a sales cycle
mes products delivered communica
on with supplier
rep or call centre programs in place
e.g CPFR or VMI
Supplier relaonship measures

FIGURE 5: Supplier relationship (respondents’ rating in %).

http://www.jtscm.co.za Open Access

 Page 9 of 11 Original Research

Collaboration programmes: Respondents were asked to To a lesser extent, respondents indicated that there was store
rate, on a four-point quality standard scale, collaboration loyalty from customers. Forty-five per cent of respondents
programmes in place with suppliers, that is, collaborative indicated that customers would postpone the purchase
planning, forecasting and replenishment (CPFR) or vendor- until the product was available. In addition, respondents
managed inventory (VMI). Figure 5 illustrates that more indicated that customer retention decreased when customers
than 86% of respondents rated a poor- to low-quality were faced with stock-outs. In terms of the likelihood of
standard of collaboration. These results suggest that the lack customers purchasing another brand when faced with a
of collaboration to create synchronisation could result in stock-out, 59% of the respondents agreed with this statement.
stock-outs at the retail level. CPFR programmes enable This is contradictory to the brand loyalty theory. This
retailers to leverage synergies with their suppliers. appears to be in the favour of the retailer; however, the
Achieving this requires sharing information on customer literature indicates that if a consumer is faced with
demand with suppliers to improve visibility and continuous stock-outs in a store they will eventually become
synchronisation. Information sharing allows a supplier or less store-loyal. Fifty-five per cent of respondents agreed
retailer to be more agile or flexible in responding to changing that, when faced with stock-out, customers would not return
customer demand. These collaboration programmes also to the store for future purchases.
assist the retailer and supplier in experiencing a long-term
economically sustainable partnership.
Conclusion and recommendations
The impacts of inefficient business processes The aim of this study was to investigate the impact of
and practices on business performance business management practices and processes on demand
variability and stock-out situations in warehouse retail liquor
Respondents were asked to rate, on a four-point scale, what
outlets in Johannesburg. Twenty-two high sales volume
customers are likely to do when faced with a stock-out in
warehouse retail outlets were surveyed. Based on the
their retail outlet. Figure 6 illustrates that 82% (medium to
conceptual framework and the literature review, there were
high adherence) of customers would tend to purchase a
competitor product during a stock-out. This is not consistent three groups of findings: variability, synchronisation and
with the finding by Van Woensel et al. (2007) who indicated communication. Demand forecasting methodologies, forecast
that the immediacy effect causes consumers to shop tracking and inventory management techniques used by the
elsewhere. The product offering in this study is a category retail outlets were investigated to determine the association
that is non-perishable or has a longer shelf-life product and between demand variability and stock-outs.
customers do not appear to have the immediacy effect
suggested in the literature. Respondents indicated that The results revealed that the retailers based customer demand
customers would tend to shop at the next closest store forecasting on judgement, naïve or simple moving average
stocking their brand. This correlates with the effect of brand forecast methods. As a result, during periods of new products
loyalty and the customer behaviour on purchasing patterns entering into the market, promotional activity and seasonal
(Corsten & Gruen 2003; Peckham 1963; Sloot et al. 2002; Van peaks, customer demand was difficult to predict. This
Woensel et al. 2007). This implies that stock-outs will highlights that skills or capabilities for advanced data
negatively affect sales and competitive advantage in the modelling were limited. The use of poor forecasting methods
alcohol retail industry. resulted in the transfer of inaccurate customer demand

No

60 Low

54 55 Medium
High
50

41 41 41
40
36
Percentage (%)

36
32
30 27
23 23
20 18 18
14 13 13
10
10
5
0 0
0
Purchase another Wait unl products Next purchase order Purchase from Drop in retenon
brand arrive increases competors rates of customers

Customer reacons to stock-outs

FIGURE 6: Customer reactions to stock-outs (respondents’ rating in %).

http://www.jtscm.co.za Open Access

 Page 10 of 11 Original Research

signals to suppliers and stock-outs. This is consistent with the also known as the Holt–Winters method to forecast basic
beer game, as poor order signals through the supply chain seasonal demand.
results in poor matching demand of with supply. The
respondents indicated that they usually tracked customer The research study implemented a model of assessment of
demand data by using excel spread sheets and computerised business management processes and practices that cause
systems. There was, however, no evidence of data analysis or stock-outs based on attribute, opinion and behavioural data
use of forecast tracking methodologies. Respondents variables. Future research can implement this model in other
indicated that they based inventory decisions on simple rules fast moving consumer goods (FMCG) industrial sub-sectors.
and standards set by management. There was no consideration Further research investigating in the field of retail stock-outs
of advanced data modelling, including variability, lead time could quantify the real value of a stock-out in businesses. The
or a targeted customer service level, to achieve optimal above research opportunities can be explored as a case study
inventory levels. Instead, safety stock modelling usually approach or a generalisation across many businesses in
involved holding a 2-week stock cover for all products FMCG industries.
throughout the year, which is prone to stock-outs, particularly
during periods of unexpectedly high demand. In conclusion, this study has demonstrated that three focus
areas, namely, demand forecasting methods, collaboration
It was apparent from the study that organisations adhered to with suppliers, and training of employees, will assist in
store operations regarding shelving and merchandising and closing the gap of poor management processes and practices
these were therefore not a hindrance for the retailer in and reduce stock-outs for retail liquor outlets in Johannesburg.
fulfilling customer demand. However, product breakages
were common, implying that product data inaccuracies and Acknowledgements
stock-outs were highly probable, suggesting limited employee
capabilities regarding product handling. Respondents
Competing interests
reported a lack of stock or sales information sharing between The authors declare that they have no financial or personal
the retailers and suppliers, with no formal collaboration relationships that may have inappropriately influenced them
programmes in place. This indicates that suppliers lack the in writing this article.
flexibility to react to the retailers’ requirements. The results
revealed that stock-outs had a medium to high negative Authors’ contributions
impact on business performance. The respondents revealed
that frequent stock-outs resulted in a decline in sales, All three authors contributed equally to the writing of this
customer retention and competitive advantage. article.

One of the limitations of this study is that results were based References
on the opinions and behaviours of respondents and may thus
Aastrup, J. & Kotzab, H., 2009, ‘Analysing out-of-stock in independent grocery stores:
present a level of bias, particularly by respondents describing An empirical study’, International Journal of Retail & Distribution Management
37(9), 765–789. https://doi.org/10.1108/09590550910975817
adequate competencies when interviewed. A key
Anupindi, R., Chopra, S., Deshmukh, S., Mieghem, J. & Zemel, E., 2012, Managing
recommendation is that, in cases where retail outlets business process flows, 3rd edn., Pearson Education, Upper Saddle River, NJ.
experience stock-outs as a result of trends and seasonality, Atrill, P. & McLaney, E., 2008, Accounting and finance for non-specialists, 6th edn.,
Prentice Hall, London.
management should consider other methods of forecasting
Baird, N. & Rosenblum, P., 2010, Defining twenty-first century merchandising,
customer demand. It is also recommended that strategic Benchmark Report: September 2010, Retail Systems Research, Miami, FL.
partnerships be established with suppliers through the Bon, A.T. & Leng, C.Y., 2009, ‘The fundamental on demand forecasting in inventory
sharing of sales and stock information to improve management’, Australian Journal of Basic and Applied Sciences 3(4), 3937–3943.
Chisadza, C., Dlamini, J., Gupta, R. & Modise, M., 2013, The impact of oil shocks on the
synchronisation. Synchronisation will facilitate supplier South African economy, viewed 02 August 2013, from http://web.up.ac.za/
flexibility to react quickly to retailers’ needs and ensure the sitefiles/file/40/677/WP_2013_11.pdf
maintenance of customer service quality. Corsten, D. & Gruen, T., 2003, ‘Desperately seeking shelf availability: An examination
of the extent, the causes, and the efforts to address retail out-of-stocks’,
International Journal of Retail & Distribution Management 31(12), 605–617.
https://doi.org/10.1108/09590550310507731
A further limitation is the small sample size. Although the
Coyle, J., Langley, C., Novack, R. & Gibson, B., 2012, Managing supply chains: A
sample selected represented a significant portion of the top logistics approach, 9th edn., McGraw-Hill, South-Western, Canada.
50% contributors to the sector’s sales, the sample was small Croxton, K.L., Lambert, S.J., García-Dastugue, D.M. & Rogers, D.S., 2002, ‘The demand
management process’, The Journal of Logistics Management 13(2), 51–66.
and the study was exploratory. It is recommended that the https://doi.org/10.1108/09574090210806423
study should be expanded across all 917 warehouse retail Fliedner, G., 2003, ‘CPFR: An emerging supply chain tool’, Industrial Management &
liquor outlets, as well as other outlets, to determine the extent Data Systems 103(1), 14–21. https://doi.org/10.1108/02635570310456850
Forrester, J., 1961, Industrial dynamics, MIT Press, Cambridge, MA.
of the problem. It is expected that this may show a larger
Frank, C., Garg, A., Raheja, A. & Sztandera, L., 2003, ‘Forecasting women’s apparel
problem than reported in this study, as sophistication in sales using mathematical modelling’, International Journal of Clothing Science and
forecasting and ordering techniques is expected to be lower Technology 15(2), 107–125. https://doi.org/10.1108/09556220310470097

in smaller outlets. It is, however, worth noting that the small Ghosh, A. & Fedorowicz, J., 2008, ‘The role of trust in supply chain governance’,
Business Process Management Journal 14(4), 453–470. https://doi.
sample size implies that the results are not generalisable to org/10.1108/14637150810888019
the population. Respondents could consider using simpler Goodwin, J.S. & Franklin, S.G., 1994, ‘The beer distribution game: Using simulation to
teach systems thinking’, Journal of Management Development 13(8), 7–15.
methods for forecasting, such as triple exponential smoothing, https://doi.org/10.1108/02621719410071937

http://www.jtscm.co.za Open Access

 Page 11 of 11 Original Research

Grant, D. & Fernie, J., 2008, ‘Exploring out-of-stock and on-shelf availability in non- Pearson, G., 2011, Out-of-stock losses cost retailers billions, viewed 12 July 2012, from
grocery, high street retailing’, International Journal of Retail & Distribution http://www.bizcommunity.com/Article/196/168/67836.html
Management 36(9), 661–672. https://doi.org/10.1108/09590550810883496
Peckham, J., 1963, ‘The consumer speaks’, Journal of Marketing 27(4), 21–26. https://
Herrick, C. & Parnell, S., 2013, Is alcohol threatening health and development in doi.org/10.2307/1248641
South Africa?, viewed 02 August 2013, from http://www.c3health.org/wp-
content/uploads/2013/07/Is-alcohol-threatening-health-and-development-in- Pramatari, K. & Miliotis, P., 2008, ‘The impact of collaborative store ordering on shelf
South-Africa.pdf availability’, Supply Chain Management: An International Journal 13(1), 49–61.
https://doi.org/10.1108/13598540810850319
Jeffery, M.M., Butler, R.J. & Malone, L.C., 2008, ‘Determining a cost-effective customer
service level’, Supply Chain Management: An International Journal 13(3), PWC, 2012, South African retail and consumer products outlook, viewed 06 July 2013,
225–232. https://doi.org/10.1108/13598540810871262 from http://www.pwc.co.za/en_ZA/za/assets/pdf/retail-and-consumer-products-
outlook-2012-2016.pdf
Jones, G. & George, J., 2009, Contemporary management, 6th edn., McGraw-Hill
Irwin, New York. Reuters, 2013, SAs retail sales growth slows, viewed 07 July 2013, from http://www.
fin24.com/Economy/SAs-retail-sales-growth-slows-20130515
Kalchschmidt, M., Verganti, R. & Zotteri, G., 2006, ‘Forecasting demand from
heterogeneous customers’, International Journal of Operations & Production Rungtusanatham, M., Rabinovich, E., Ashenbaum, B. & Wallin, C., 2007, ‘Vendor-
Management 26(6), 619–638. https://doi.org/10.1108/01443570610666975 owned inventory management arrangements in retail: An agency theory
perspective’, Journal of Business Logistics 28(1), 111–135. https://doi.
Lee, H., Padmanabhan, V. & Whang, S., 1997. ‘Information distortion in a supply chain: org/10.1002/j.2158-1592.2007.tb00234.x
The bullwhip effect’, Management Science 43(4), 546–558. https://doi.
org/10.1287/mnsc.43.4.546 Sahay, B., 2003. ‘Supply chain collaboration: The key to value creation’, Work Study
52(2), 76–83. https://doi.org/10.1108/00438020310462872
Maswanganyi, N., 2013, Slump in SA retail sales growth fans fears that household
spending will slow, viewed 06 July 2013, from http://www.bdlive.co.za/ Sahay, B. & Mohan, R., 2003, ‘Supply chain management practices in Indian industry’,
economy/2013/05/16/slumpin-sa-retail-sales-growth-fans-fears-that-household- International Journal of Physical Distribution & Logistics Management 33(7),
spending-will-slow 582–606. https://doi.org/10.1108/09600030310499277
Miranda, M. & Jegasothy, K., 2008, ‘Malaysian grocery shoppers’ behavioural response Salmi, L. & Holmström, J., 2004, ‘Monitoring new product introductions with sell-
to stock-outs’, Asia Pacific Journal of Marketing and Logistics 20(4), 396–412. through data from channel partners’, Supply Chain Management: An International
https://doi.org/10.1108/13555850810909722 Journal 9(3), 209–212. https://doi.org/10.1108/13598540410544890
McNaghten, M. & Passino, C., 2011, Understanding customer retention in the retail Simatupang, T., Sandroto, I. & Lubis, S., 2004, ‘Supply chain coordination in a fashion
industry, viewed 13 October 2012, from http://www.deloitte.com/assets/Dcom- firm’, Supply Chain Management: An International Journal 9(3), 256–268. https://
Peru/Local%20Assets/Documents/CB_understanding_customer_retention.pdf doi.org/10.1108/13598540410544953
Nakano, M. & Oji, N., 2012, ‘The transition from a judgemental an integrative method Sloot, L., Verhoef, P. & Franses, P., 2002, The impact of brand and category
in demand forecasting: A case study of a Japanese company’, International Journal characteristics on consumer stock –out reactions, viewed 05 August 2013, from
of Operations and Production Management 32(4), 386–397. https://doi. http://repub.eur.nl/res/pub/257/erimrs20021125174407.pdf
org/10.1108/01443571211223068
Van Woensel, T., Van Donselaar, K., Broekmeulen, R. & Fransoo, J., 2007, ‘Consumer
Onwuegbuzie, A. & Leech, N., 2005, ‘On becoming a pragmatic researcher: The responses to shelf out -of-stocks of perishable products’, International Journal of
importance of combining quantitative and qualitative research methodologies’, Physical Distribution & Logistics Management 37(9), 704–718. https://doi.
International Journal of Social Research Methodology 8(5), 375–387. https://doi. org/10.1108/09600030710840822
org/10.1080/13645570500402447
Wisner, J., Tan, K. & Leong, G., 2012, Supply chain management: A balanced approach,
Onwuegbuzie, A.J., & Teddlie, C., 2003, ‘A framework for analyzing data in mixed 3rd edn., South-Western Cengage Learning, Mason, Canada.
methods research’, in A. Tashakkori & C. Teddlie (eds.), Handbook of mixed
methods in social and behavioral research, pp. 351–383, Sage, Thousand Oaks, CA. Yin, R., 2014, Case study research design and methods, 5th edn., Sage, Thousand
Oaks, CA.
Pawlak, M. & Malyszek, E., 2008, ‘A local collaboration as the most successful co-
ordination scenario in the supply chain’, Industrial management and data systems Zikmund, W., 2003, Business research methods, 7th edn., Thomson, South-Western,
108(1), 22–42. https://doi.org/10.1108/02635570810844061 OH.

http://www.jtscm.co.za Open Access