Demand Forecasting &

Order Fulfilment in Supply

Chains
 Learning Objectives
 To understand the types of forecasts that may be needed.
 To understand how collaboration among trading partners
will help the overall forecasting and demand
management process.
 To identify the key steps in order fulfillment process and
understand how effective order management can create
value for a firm and its customers.
 Realise the meaning of customer service and understand
its importance to logistics and supply chain management.
Traditional Forecasting – Demand Forecasting
 A major component of demand management is forecasting
 the amount of product that will be purchased, when it will
be purchased, where it will be purchased by consumers or
end users.
 Although many techniques are available to forecast demand,
they are all invariably wrong.
 The key to successful forecasting is to minimise the error
between actual demand and forecasted demand.
 There are many factors in market that will change actual
demand as against forecasted demand.
 However, forecasts are essential for marketing and operations
to set goals and develop execution strategies (S&OP process).
 Demand Forecasting
 Demand forecasts form the basis of all supply chain planning.
 Consider the push/pull processes in a supply chain.
 All push processes in supply chain are performed in
anticipation of customer demand.
 whereas pull processes are performed in response to customer
demand.
 For push processes, supply chain planners must plan the level
of activity (production, transportation, etc.).
 For pull processes, a manager must plan the level of available
capacity and inventory, but not the actual amount to be
executed.
 Demand Forecasting
 In both instances, the first step the supply chain planners

must take is to forecast what customer demand will be.

 Dell orders PC components in anticipation of customer orders,
whereas it performs assembly in response to customer orders.

 Dell uses a forecast of future demand to determine the quantity

of components to have on hand (a push process) and to
determine the capacity needed in its plants (for pull
production).
 Factors Affecting Demand
 Two types of demands exist: (i) Independent and (ii)
Dependent .
 For some items, their demand is independent (bicycles).
This is the demand created directly by the final users
(consumers).
 Dependent demand is influenced by the demand for
independent items.
 Thedemand for bicycle tires is dependent on demand for
bicycles.
 Most forecasts focus on independent demand.
 Bicyclemanufacturer focuses on demand for bicycles and
estimates the demand for tires, etc.,
 Characteristics of Forecasts
 Companies and supply chain managers need to be aware of

the following characteristics of forecasts:

 Forecasts are always wrong and should thus include both the
expected value of the forecast and a measure of forecast error.
 Forecast error (demand uncertainty) must be a key input into
most supply chain decisions. Unfortunately, most firms do not
maintain any estimate of forecast error.
 Long-term forecasts (larger standard deviation) are usually less
accurate than short-term forecasts (smaller standard deviation).
 Characteristics of Forecasts
 Companies and supply chain managers need to be aware of

the following characteristics of forecasts:

 Aggregate forecasts (smaller standard deviation) are usually
more accurate than disaggregate forecasts (larger standard
deviation) (forecasts at a company level is more accurate than
forecasts at a product level).
 The father up a supply chain a company is (farther from the
consumer), the greater is the distortion of information it
receives (relate it to Bullwhip effect).
 As a result, the farther up the supply chain an enterprise is, the
largest will be the forecast error.
 Traditional Forecasting – Demand Forecasting
 The demand for independent demand is aka base demand
(normal demand).
 All demands are subject to certain ‘fluctuations’.
 One type of fluctuation is random variation. This can’t be
anticipated (that’s why safety inventory to avoid stockout).
 Another type is called trend. Refers to gradual increase or
decrease in demand over time (increase in demand for
iPad and reduction in demand for DVDs/VCRs).
 Another type is called seasonal variation. Repeated during
a year for most organisations (sun cream, winter cream).
 Demand fluctuations can also be caused by business
cycles (growth, stagnant, decline).

This fluctuation occurs for more than a year.
 Traditional Forecasting – Demand Forecasting
 It was already suggested that all forecasts will be wrong.
 Most forecasts are higher/lower than actual demand.
 The key to successful forecasting is to minimize error
between forecasted demand and actual demand.
 Forecasting demand is highly scientific art and various
quantitative techniques exist to manipulate historical data
to predict future.
 However, an assumption is made here that past will be
repeated in the future.
 This is normally NOT the case.
 Sales and Operations Planning
 Historically many firms developed several functional
forecasts for same products for same time period.
 It is possible for a manufacturer to have financial
forecast, manufacturing forecast, marketing forecast,
distribution forecast.
 The problem with multiple functional forecasting is that
most of the times these functional forecasts dod not
agree.
 E.g: Marketing would forecast higher demand that
neither manufacturing nor distribution could execute.
 Finance forecasts would be higher than marketing
would be able to meet.
 Sales and Operations Planning
 It is necessary for a firm to arrive at an agreeable
forecasts internally that can be used execute.
 This process is called ‘sales and operations planning
consensus.’
 This involves a five-step process

Step 1 Step 2 Step 3

Run Sales Demand Supply Planning
Forecast Reports Planning Phase Phase

Step 5 Step 4
Executive S&OP Pre-S&OP
Meeting Meeting
 Sales and Operations Planning
 Step 1 - Sales Forecast Reports:
 Requires development of statistical forecasts for the future
sales. Firms use many techniques to develop sales forecast
reports.
 Step 2 – Demand Planning Phase:
 The sales and marketing departments to review the
forecast and make adjustments based on promotion of
existing products, the introductions of new products, or
the elimination of products.
 The revised forecasts is stated in terms of both units
(operations department) and pound/dollar since (finance
department).
 Sales and Operations Planning
 Step 3 - Supply Planning Phase:
 Requires operations (manufacturing, warehousing,
transportation, etc.,) to analyse the sales forecast to
determine if existing capacity is adequate to handle
forecasted volumes.
 Existing capacity might be adequate if demand is stable.
However, if ‘promotions’ are introduced, demand will
spike (capacity may not be sufficient).

Two options available to manage capacity constraints.

Promotions could be curtailed to bring demand down (lost
sales/revenue/profit).

Additional manufacturing capacity could be generated
internally or externally (contract manufacturing/
outsourcing, etc.,) with additional costs. This needs to be
repeated for other operations as well.
 Sales and Operations Planning
 Step 4 – Pre-S&OP Meeting:
 This process involves individuals from sales, marketing,
operations, and finance.
 Initial forecasts will be reviewed along with any capacity
issues that emerged in step 3.
 Attempts made to resolve capacity issues by attempting
to balance supply and demand.
 Alternative scenarios are developed to be considered at
executive S&OP meeting (Step 5).
 These scenarios will identify potential lost sales and
increased costs associated with balancing of supply and
demand.
 Sales and Operations Planning
 Step 5 – Executive S&OP Meeting:
 Final decisions are made regarding sales forecasts and
capacity issues.
 Executives from various functional areas agree on a
forecast and convert it into operating plan for firm.
 Consensus among functional areas is critical here.
 Decisions regarding trade-offs between revenue and
costs are made.
 The final decision is approved in this meeting for
execution.
 S&OP to CPRF
Problems in Traditional Forecasting

 The S&OP process describes how firms arrive at a

consensus on forecasts internally.

 The next step is for the members of a supply chain to

agree upon a consensus forecast.

 Problems in Traditional Forecasting
 Many initiatives attempted to create efficiency and
effectiveness through the integration of supply chain
activities and processes
 QR (quick response)
 VMI (vendor managed inventory)
 CRP (continuous replenishment planning),
 ECR (efficient customer response).
 EDI (electronic data interchange)
 However, they were all found to be inefficient as there
was no incentive for collaborative planning in these
arrangements.
 Problems in Traditional Forecasting
 They fell short of expectations in integrating supply

chain activities and processes with participants.

 An important initiative aimed at achieving true supply

chain integration is CPFR.

 Collaborative Planning, Forecasting, and
Replenishment (CPFR)

 CPFR is recognised as a breakthrough business model

for planning, forecasting and replenishment.
 In this method, knowledge and information that exists
internally and externally is brought together into a single,
more accurate, forecast that has the support of the entire
supply chain.
 Manufacturers, retailers, transport providers, distributors
use internet-based technologies to collaborate from
operational planning through execution.
 Collaborative Planning, Forecasting, and
Replenishment (CPFR)
 Previously, manufacturers and retailers have had 20 or
more forecasts between them for single product.
 Eachmore or less accurate, all trying to predict consumer
behaviour.
 CPFR simplifies and streamlines overall demand
planning.
 CPFR allows trading partners to agree on a single
forecast for an item where each partner translates this
forecast into a single execution plan.
 This replaces traditional forecasting methods, where
forecasts are made throughout the chain (not connected
to primary demand – Mentzer & Moon – ‘Islands of
analysis’).
 Collaborative Planning, Forecasting, and
Replenishment (CPFR)
 Previously, manufacturers and retailers have had 20 or
more forecasts between them for single product.
 Eachmore or less accurate, all trying to predict consumer
behaviour.
 There were no co-operational aspect to traditional
forecasting.
 Developed by Wal-Mart and Warner-Lambert in 1995
(Listerine product).
 These two firms, in addition to collaborating to
rationalise inventories and out-of-stock occurrences, also
collaborated to increase the forecast accuracy.
 Collaborative Planning, Forecasting, and
Replenishment (CPFR)

 Three month pilot produced significant results and

improvements for both parties thereby resulting in the

adoption of this approach.
 The CPFR Model

Larry Smith, West Marine: A CPFR Success Story, Supply Chain

Management Review (March, 2006).
 CPFR Model
 The model is a sequence of several business processes
that include the consumer, retailer and manufacturer.
 The four major processes are (i) Strategy and Planning;
(ii) Demand and Supply Management; (iii) Execution
and (iv) Analysis.
 CPRF emphasises a sharing of consumer purchasing data
(point of sale data) as well as forecasts at retail among
and between partners to help manage supply chain
activities.
 From these data, the manufacturer can analyse his ability
to meet forecasted demand.
 CPFR Model
 If he can’t meet the demand, a collaborative effort is
undertaken between the retailer and manufacturer to
arrive at a mutually agreed-upon forecast from which
execution plans are developed.
 The strength of CPRF is that it provides a single forecast.
 The trading partners can develop manufacturing
strategies, replenish strategies and merchandising
strategies based on this single forecast.
 CPFR Model

 CPFR emphasises a sharing of consumer purchasing data

among and between supply chain partners.

 In this regard, CPFR creates a direct link between the

consumer and the supply chain.

 The plan and the forecast are entered by suppliers and

buyers into an Internet accessible system.

 Within established parameters, any of the participating

partners are empowered to change the forecast.

 CPFR Model
 The results of CPFR initiatives are impressive (Wal-Mart,
Sara Lee reported reduction of inventory and increase in
sales).
 Collaborative planning improves the quality of the demand
signal for the entire supply chain through a constant
exchange of information from one end to the other.
 Goes beyond the traditional practice.
 Effective implementation is based on systematic
collaboration between trading partners (not the previous
approaches).
 CPFR Model
 Warner-Lambert’s service levels increased from 87% to
98%, while the lead times to deliver the product
decreased from 21 to 11 days.
 The partnership also increased Listerene sales by $8.5
million over the test period (Hill, 1999).
 The success prompted the Voluntary Interindustry
Commerce Standards (VICS) to set up guidelines for
synchronising business processes, forecasts, and
replenishments, now formalised as CPFR.
 The central theme of the CPFR guidelines was to align
processes and standardise technologies to share forecast
and other planning information securely, simultaneously,
globally and in real-time.
 Fulfillment Models
 Early part of this lecture focused on planning and
forecasting aspects of demand management.
 Once these aspects are agreed upon, the plan needs to be
executed (called fulfillment process).
 Various distribution channels and strategies are used to
deliver on the demand forecast.
 A channel consists of one or more organisations who
participate in the flow of goods, services, information
and finances from the point of production to point of
consumption.
 A channel consists of both logistics channel and a
marketing channel.
 Channels of Distribution
 The logistics channel refers to the means by which the
products flow from point of production to point of
consumption.
 The marketing channel refers to the means by which the
necessary transactional elements are managed (such as
customer orders, billing, accounts receivable, etc.).
 The logistics and marketing channels need to work
together in fulfilling customer orders.
 Numerous channels are responsible for delivering
products to consumers.
 Channels of Distribution
 It is important to note that channel structure involves

elements of fixed costs and variable costs.

 Distribution centres, stores, etc. – Fixed costs
 Transportation, packaging, etc. – variable costs
 Organisations tend to choose distribution channel that

reduces fixed costs and increases variable costs (while

managing the fulfillment process).
Fulfilment Models
 Direct-to-Customer Fulfilment
 Integrated Fulfillment:
 Many firms operate both bricks-and-mortar and click-and-
mortar (Tesco, Sainsbury, M&S, etc.).

i.e. Retailers have both retail stores & internet sites.
 In integrated fulfillment strategy, a retailer operates one
distribution network to service both channels.
 Direct-to-Customer Fulfilment
 Dedicated Fulfillment:
 Another option for the retailer to desire to have both a
store and an internet presence.
 This option achieves the same delivery goals, but with two
separate distribution networks.
 This option eliminates some disadvantages of previous
distribution network. However, one channel,
unfortunately, duplicates another channel.
 Direct-to-Customer Fulfilment
 Outsourced fulfillment:
 While the previous two fulfillment options assume that the
retailer would perform fulfillment, outsourced fulfillment
assumes that another outside firm will perform the
fulfillment.
 Direct-to-Customer Fulfilment
 Drop-Shipped Fulfillment:
 The manufacturer delivers the product direct to the
retailer’s stores, by passing the retailer’s distribution
network. A major advantage is reduction of inventory in
distribution networks.
 Direct-to-Customer Fulfilment
 Store Fulfillment:
 For a retailer with both store & internet presence store
fulfillment offers several opportunities.
 The order is placed through internet site. It is sent to the
nearest retail store, where it is picked and put aside for the
customer to pick up.
 Conclusions
 Although many forecasts are made throughout the supply
chain, the forecast of primary demand from end user or
consumer will be the most important.
 The S&OP process has gained much attention as it serves
the purpose of allowing a firm to operate from a single
forecast.
 CPFR is a method to allow trading partners to
collaboratively develop and agree upon a forecast.
 A number of distribution channels are available to fulfill
customer orders.