Vous êtes sur la page 1sur 66

Order Fulfillment

Richard Pibernik
Lehrstuhl fr Logistik und Quantitative Methoden

1
Remember!

Strategic Network Design

SC Master Planning
Demand
Planning APS
Planning Horizon

Material Production Distribution


Requirements Planning Planning
Planning
& Transportation Order
Purchasing Scheduling Fulfillment
Planning You are here!

Monitoring/Event Management

Materials
Management Production
Planning
Sales & Distribution ERP*
Purchasing

*Relevant modules for Supply Chain Management


2
Learning Objectives

So far, we have addressed planning problems that occur upstream


relative to the decoupling point (i.e., that are based on forecasts)
Now, we consider decisions that have to be taken downstream (i.e.
when customer orders/demand have materialized)
Once customer orders/demand are known, we have to decide how
to best fulfill them given our available supply (i.e., capacity &
inventory) this is commonly referred to as order fulfillment or
demand fulfillment
After completing this part of our course, you should
Know the basic tasks and functionalities of order fulfillment (systems)
Understand key objectives and relevant trade-offs
Know different modes of order fulfillment and when they should be employed
Have a thorough understanding of the logic of order fulfillment systems and
state-of-the-art quantitative methods
Have a good knowledge of how to cope with stock-out situations
3
Order Fulfillment

balances supply and demand in the short run

Some example (industry) problems:

Dell receives a customer order for 20 high-grade servers. For when should they promise
delivery?

An OTC pharmaceuticals company runs short on stock. Which customers should receive
their deliveries first, which customers have to wait?

A hard disk manufacturer runs short but has the option to upgrade or expedite certain
orders. Should the manufacturer do so?

Alcatel-Lucent has promised certain customers a specific service level for DSL
equipment. How much capacity and inventory should they reserve for these customers?

A paper manufacturer wants to incentivize customers to order earlier. What should they
offer?

4
Important to understand: Decoupling Point

Fleischmann/Meyr, 2003

Order fulfillment occurs downstream (i.e., on the right hand side)


5
of the decoupling point
Order Management/Order Fulfillment

The order fulfillment process includes all


relevant activities from placing an order by the
customer to its fulfillment/completion Order Entry

Objectives Order Promising


Enable easy to use (electronic) order placement and
order entry
Short response time to customer orders and
requests, Scheduling
Promise reliable due dates and ensure on time (Production)
delivery
Scheduling
Optimal allocation of orders to resources (Shipping)
(administrative resources, inventory,
manufacturing, distribution and transportation Execution &
resources) Monitoring

6
Different Terms

Note: different terms are being used in industry and academia, which
are not precisely defined and are often overlapping:

Order fulfillment

Demand fulfillment

Order management

Order promising

Available to promise (ATP)/Capable to promise (CTP)/Profitable to Promise


(PTP)

Due date Quoting

7
The Fulfillment Process

Check availablitiy
Customer order Order has been
received confirmed

XOR

Order entry Execution and


Reuested Quantity not monitoring
quantity available on
is available time
Order has been Determine alternative
entered strategies Order fulfilled

Alternative
strategies
determined
XOR

Confirm order and


delivery dates

8
Tasks and Objectives

Tasks of order fulfillment systems:


Receive/handle orders coming in through multiple channels
Quote due-dates/deliver dates (order promising)
Reserve capacity/schedule order
Order execution and monitoring/rescheduling

Objectives:
Enable easy to use electronic order placement and order entry
Short response time to customer orders and requests
Ensure on time delivery
Optimal allocation of incoming orders to inventory, manufacturing, distribution and
transportation resources

To which criterion/criteria does the term optimal refer to?

9
Receiving Orders

Today, systems have to be able to support order entry through:


conventional channels (fax, e-mail)
phone (call center support)
website
Electronic Data Interchange

For a number of companies, so-called online configurators are essential to their business
model (ATO/MTO)
Most prominent: Dell, started in 1996
Features include
compatibility constraints
quoting
approximate delivery time

Although still relatively young, these functions are now mature; state of the art systems are
fairly easy to implement and run stable

10
More Interesting: Order Promising

Relevant questions:

When can we fulfill an order?

Are we meeting the customers requirements in terms of delivery times?

What can we do if we cannot meet the customers requirements? Any alternatives?

If we have alternatives, how much do they cost us? Is the customer worth the additional
cost?

In upcoming stock-out situations, which customer orders are fulfilled, which orders are
rejected or have to wait? What are short and long term consequences?

Order promising links sales and CRM to core SCM functions (manufacturing,
inventory management, distribution)

11
Order Promising

Lets have a closer look at the order promising process described previously

Check availablitiy

XOR

Reuested Quantity not


quantity available on
is available time
Determine alternative
strategies

Alternative
strategies
determined
XOR

Confirm order and


delivery dates

12
Conventional ATP

Conventional Available to Promise (ATP) implemented in ERP


Provides information on uncommitted portion of inventory on hand + master schedule
Simple bookkeeping function:

ERP module: SD

ERP module: PP

ERP module: MM

t=1 t=2 t=3


Initial inventory 100
Planned receipts 300 500 500
Committed quantities 200 400 200
ATP 200 300 600

13
Order Promising Make to Stock

Slightly advanced order promising includes:

Simultaneous allocation of orders to ATP quantity and due date quoting

When can an order of size be fulfilled?

ATP

Most commonly, due date quoting against available to promise inventory is called ATP
Easy to implement standard function

14
Order Promising Assemble to Order

ATP Component A
ATO order promising includes:

Simultaneous allocation of orders to component


ATP quantities and due date quoting
t
ATP Component B
When can an order be fulfilled if?

capacity and quantities

ATP Component C t
of A,B,C are required?

Capacity and all components can be bottleneck


Most commonly referred to as Capable to t
Promise (CTP) Assembly Capacity
Relatively easy to implement standard function

15
Order Promising Make to Order

If we want to consider a multi-stage production process, order promising gets


very complicated if bottlenecks can occur on more than one stage

If only one bottleneck, ATO logic can be applied

We will look at MTO order promising later

16
Real-Time Order Promising (1)

Real-time order promising


For every incoming order, delivery dates are immediately quoted
Customer instantly receives information about availability and expected delivery date

Inventory of component 1
(q16 , d16 )
mM
invm,t
Incoming Orders

d16 Allocation of resources;


t due date quoting
Capacity

capt

3
(q15 , d15 ); d15
Order book
t
(q24 , d2 ); d2 (with quoted
due dates)
(q12 , d1 ); d1
17
Real-Time Order Promising (2)

Real-time Capable to Promise (CTP) algorithm:


Quantity of product p that can be produced in time
d interval [tb+1,d]
ctp p ,d ctp p ,t ,d
t t b 1 Inventory component 1,,M in period d
Capacity in period d

inv 1,d inv M ,d capd
ctp p ,d ,d min

1, p
,...,
M ,p
,


p
for m ,p 0

Resource coefficients for product p

inv mnet,d inv m ,d ctpp ,d ,d m , p

net

inv 1,t 1 inv Mnet,t 1 inv 1,t inv M ,t capt
ctp p ,t ,d min ,..., , ,..., ,
p
1, p M ,p 1, p M ,p
t d 1,..., t b 1
net
net

inv m ,t min inv m ,t , inv m ,t 1 ctp p ,t ,d m ,p
m M ,t d 1,...,t b
18
1
Batch Order Promising (1)

First we collect all incoming orders during a batching-interval, e.g. one day
At the end of the batching-interval, orders are allocated to inventory/
capacity and due dates are determined

Inventory of component Order book


3
mM (q15 , d15 ) (end of batching-
invm,t interval)
(q24 , d 2 )
(q12 , d1 )

t d1 , d2 ,..., d15
Capacity Allocation of resources;
due date quoting
capt

t 3
(q15 , d15 ); d15 Order book
(with quoted
(q24 , d 2 ); d2 due dates)
(q12 , d1 ); d1 19
Order Promising Real Time vs. Batch

It should be noted that there are very few true real-time ATP systems operating today;
most systems that give an immediate response (including most web-based retail sites)
produce an initial soft promise, run a batch module later and then produce a hard
promise.
(Ball et al., 2002)

What are the advantages/disadvantages of real-time and batch order promising?

When employing batch order promising, the length of the batching-interval is a decision
variable
We will look at models supporting batch order promising!

20
Batch Order Promising (2)

Alternative mechanisms:
sequential (e.g. first come first served, highest priority first)
optimization models

Lets look at a simple optimization model

21
Batch Order Promising Model Notation

Additional variables
x ip,t production quantity product p for order i in period t
x ip,t,cum cumulated production quantity for order i in period t
x ip,t quantity shipped for order i in period t
1 if order i is shipped in t
y i ,t
0 else
1 if order i is promised
zi
0 else

Additional parameters:
i profit margin order i
c i (y i ,t ) penalty cost if order cannot be fulfilled

22
Model Formulation (1)

23
Model Formulation (2)

xip,t m , p

24
Active vs. Passive Order Promising

Passive order promising


Utilizes information about available inventory/capacity for due date quoting
Returns due dates for execution to manufacturing, warehouse management, and
distribution systems
Does not generate detailed schedules for order execution (e.g. manufacturing
schedules)

Active order promising


Utilizes information about available inventory/capacity for due date quoting
Simultaneously determines schedules for execution (e.g. manufacturing schedules)

When are active/passive systems feasible?


What are the advantages/disadvantages?

25
Classification of Order
Promising Types and Methods

Availability Level

Finished goods Resources

Real- disposition
dispositionrules
rules disposition
dispositionrules
rules
time RT/FG/A RT/FG/P RT/R/A RT/R/P
Response- (RT)
time

?? ?? B/R/P
Batch
(B)
B/FG/A B/FG/P B/R/A
optimization models
optimization models
Aktive (A) Passive (P) Active (A) Passive (P)

Execution Level

Simultaneous generation Determiniation of due dates


of due dates and first, then scheduling
schedules

26
Search rules in case of shortage

Search rules for supply alternatives:


(Rule-based search, Multi-dimensional search), e.g. SAP-APO:

CTP 1 substitute products


4 2 same product at other
locations

3 substitute products at
other locations
4 CTP: create new
2 Locations supply order

1 3

Products
27
Order Promising with substitute products

In certain cases substitute products can be delivered instead of the product,


originally ordered by the customer

Depends on the availability of a product acceptable to the customer

Customer will only accept the substitute if it provides at least the same utility as
the original, and if its use is not limited, e.g. by technical restrictions

Necessary to determine additional fulfillment cost. Do we really want to carry


these?

28
Order Promising with multiple locations

If order cannot be fulfilled with the finished goods or capacity at a certain


location, order can be shifted to other locations

Order promising is then applied to a distribution or manufacturing network


rather than only to a single location

Multi-location order promising should take different manufacturing and


transportation lead times and costs into account

29
Order Promising with partial deliveries

If ordered quantity is not available within requested delivery time window, order
can be fulfilled with two or more partial deliveries

First partial delivery should be scheduled within the given time window

Customer must agree to partial deliveries

If partial deliveries are possible, then order promising should determine the
quantities and delivery dates for each partial delivery

Necessary to determine additional fulfillment cost. Do we really want to carry


these?

30
Assessment of Fulfillment Alternatives

The previously outlined Add-ons can be utilized in case of temporary stock-


out situations

They can be employed simultaneously and can also be combined (Note the
increase in complexity)

System should then be able to assess the fulfillment alternatives in terms of


fulfillment cost
customer satisfaction?

Critical question: How well does a system support generation and assessment
of these fulfillment alternatives?

31
Order Promising/Fulfillment in hierarchies

So far, we have mainly looked at order fulfillment for a


single location
In many companies, however, we find different geographical
(sales) regions across which product quantities are first allocated
in a hierarchical manner
The allocated quantities are the basis for order promising
In the following we will address this problem and see how this is
handled in SAP APO
Slides come from Kilger/Meyr, Chapter 9 in
Stadtler/Kilger (2010)

32
The Customer Hierarchy

To allocate supply to customers the following is required:


a model of the customer structure
aligned with the geographic dimension in demand planning
a forecast of the future customer demand

Forecast
Aggregation
World-wide
top 1400 Sales

area 1400 America Europe

400

country 1000 Germany France Italy

400 200

region 400 West East

200 200

33
The allocation process first step

Aggregate all forecast quantities to the root of the hierarchy.


This gives the total forecast of the specific product (or product
group).
Transfer total forecast to Master Planning (MP) to check which
portion of the forecast can be fulfilled.
Example: 1400 total forecast; 1200 can be fulfilled (according to
MP)

34
The allocation process second step

Allocate MP quantity top down to the leaves of the customer hierarchy


This top-down allocation is controlled by allocation rules:
rank based:
higher priorities (ranks) between customers
available quantity is allocated to higher rank up to the original forecast
helpful to support sales to specific markets
per committed:
allocation according to the forecast the customers have committed to
split appropriate to the fraction of the original forecast
support a fair share allocation according to the forecast
caution: rationing gaming may occur !
fixed split:
apply predefined split factors
independent of the individual forecast
35
Example for second allocation step

ATP
Allocation
World-wide
top 1200 (1400) Sales

rank based
rank 1 America Europe rank 2
area 1200 (1400) 400 (400) 800 (1000)

per committed

country 800 (1000) Germany France Italy

320 160
fixed split (400) (200)

region 320 (400) West East

split 0.7 split 0.3


224 (200) 96 (200)

36
Search process

General ATP-based order promising process:


search for ATP according to a set of search rules
if ATP is found
reduce accordingly and generate a quote for the order
if no ATP is found
no quote is generated and the order must be rejected or confirmed
manually at the end of the allocation planning horizon
(SC is not able to fulfill the order within the allocation planning horizon)

ATP is searched along several dimensions (as shown before)


time dimension
customer dimension
product dimension
37
Example of an ATP search procedure

Three dimensions of ATP search paths


World-wide
Sales

America Europe

2. Customer

Germany France Italy

3. Product

West East Request

1. Time

38
Search rules for example
Following search rules are applied:
(assume that ATP is on finished goods level)
1. The leaf node in the customer hierarchy, to which the customer belongs, the
product being requested by the order and the time bucket containing the
customer requested date are determined.
ATP at this point is consumed if available
2. If ATP is not sufficient, then the time dimension is searched backwards in time
for additional ATP (still for the same customer and product)
all ATP found up to a predefined number of time buckets back in time is
consumed
(consumption in earlier time buckets leeds to pre-built of an order = inventory)
3. If ATP is not sufficient, steps 1. and 2. are repeated for the next higher node
(parent node) in the customer hierarchy, then for the next higher and so on
up to the root
4. If ATP is not sufficient, steps 1. to 3. are repeated for all alternate products
that may substitute the original requested product (e.g. downgrading)
5. If ATP is not sufficient, steps 1. to 4. are repeated, but searched forward in
time, up to a predefined number of time buckets.
39
(order is made late)
ATP Consumption steps 1 & 2
Order:
300 units
from customer in East Germany
requested date in week 4

World_Wide_Sales -> Europe -> Germany -> East

Request
1. Time 300 units
70 50 60 10
week 4
weeks
1 2 3 4

1. ATP check for customer group East Germany for week 4


2. ATP check for week 3 and for week 2
(allowed to consume ATP 2 weeks backwards in time)
ATP found:
10 in week 4, 60 in week 3, 50 in week 2 = 120 in total 40
ATP Consumption step 3

3. ATP check in next higher customer node Germany


(but no ATP in Europe or World-Wide is available)
ATP found
80 in week 4, 30 in week 3, 10 in week 2 = 120 in total

2. Customer
Germany
20 10 30 80
1 2 3 4 weeks

West East
30 20 40 30 70 50 60 10 Request
300 units
1 2 3 4 weeks 1 2 3 4 weeks
week 4

41
ATP Consumption step 4

4. ATP search for alternate products


(alternates are sorted by priority)
alternate with highest priority is considered first and the same steps are
applied as for the original product
(first search back in time and second search up the customer hierarchy)
same steps to the alternate with second priority and so on

East, Product A, alternate 3


East, Product A, alternate 2
3. Product East
East, Product A, alternate 1
East
East, Product A
East

70 50 60 10 Request
1 2 3 4 weeks 300 units
week 4

42
Managing Stock-outs with
Order Fulfillment Systems

43
Contents/Objectives

One of the key criteria to evaluate order fulfillment systems is their capability to
handle temporary stock-out situations

If a company is facing a temporary stock-out situation, it will want to minimize the


overall (i.e. short and long term) negative consequences

In this session we will analyze, which negative consequences result from stock-
out situations and how they can be mitigated through the use of order fulfillment
systems

A case study from the pharmaceuticals industry will give you insight into specific
problems regarding the management of stock-out situations and will facility a
discussion about suitable models and techniques for handling these situations
effectively

44
Impact of Order Fulfillment on Profitability

Order promising is trivial if we always have enough capacity/inventory


available to fulfill customers requirements!

But,
we want systems to be tight (i.e. minimize excess inventory, maximize capacity
utilization)
we have to cope with uncertainties (supply, manufacturing, demand)

How can we handle situations in which not all customer requirements


(quantity and delivery date) can be fulfilled?

45
Impact of Order Fulfillment on Profitability

If inventory/capacity is not sufficient for fulfilling all customer requirements:

Allocation of available inventory/capacity to incoming customer orders

More appropriate: allocation of stock-outs to customer orders

Allocation has impact on short and long term profitability:


Short term: Lost sales, penalties, additional cost for expediting or upgrading
Immediate impact on profitability

Possible trade-off

Long term: Customer satisfaction, loss of goodwill, (loss of reputation)


Impact on Customer Lifetime Value, i.e. on long term
profitability

46
Traditional Inventory Theory

Stocking decisions based on:


Demand forecasts/probability distribution of (anonymous) demand
Based on aggregate data (e.g. demand in cycle time)

Suggests: Determine inventory policy that minimizes

holding cost + fixed set-up/ordering cost + penalty cost

holding cost + fixed set-up/ordering cost


Implemented!
subject to: (/) service level constraint

Determines potential for order fulfillment

47
Traditional Inventory Theory vs.
Order Fulfillment

Inventory Management Order Fulfillment


SC Planning SC Execution
Based on Forecasts/aggregate anonymous Based on detailed order information
demand
Stochastic Deterministic
Demand fill rate (represented ex ante Order fill rate based on required due dates
through /-service level) and quantities
Penalizes unfilled fraction of aggregate Penalty should depend on customer
demand (order)

Order fulfillment is especially concerned about temporary stock-outs, i.e.


materialization of 1-/1-
situations where aggregate but not disaggregate demand can be fulfilled
situations caused by SC disruptions (supply shortages, yield problems, )

48
Impact of Order Fulfillment on Profitability

Temporary stock-outs ways to Better, but not necessarily


make our customers unhappy feasible/available to all customers
Reject order
Quote unrealistic delivery date Meet customer requirements with
Quote late delivery date remaining capacity/inventory
Quote less than ordered quantity Upgrade
Partial deliveries (first delivery meets Expedite
delivery date requirements)
Partial deliveries (first delivery late)

Different short and Different impact on


long term consequences fulfillment cost
(highly customer specific!)

Has to be decided for every individual order!

49
Impact of Order Fulfillment on Profitability

What are the long term consequences of not meeting a customers requirements
in regard to a specific order?

Customers reaction:
No reaction ( CLV= CLVb - CLVa = 0)
Chooses different supplier ( CLV = CLVb)
Chooses different supplier, may return after a while ( CLV > 0)
Procures portion of products from different suppliers ( CLV > 0)

Customers reaction depends on:


Availability of alternative suppliers (not order specific)
Contractual relationship (not order specific)
Utility/need/criticality of order (i.e. impact on customer side)
Deviation from customers requirements (tardiness)
Fulfillment history

50
Objective of Order Fulfillment in
Stock-out Situations

In a temporary stock-out situation, order fulfillment should allocate


available inventory/capacity to customer orders in such a way,
that the overall negative impact is minimized

Take into account impact of individual orders and importance of customers


(represented by CLV)

Exploit all options (e.g. partial deliveries, upgrades, expediting) if economically


feasible (i.e. additional cost less than negative consequences CLV)

51
Case Data

General information:
Pharmaceuticals company, operating mainly in Europe, 1000 employees, 378 M revenue in
2003/2004
Products: Variety of prescription and over the counter drugs
Relevant products for case: over the counter pharmaceuticals (e.g. cough drops, herbal laxatives,
vitamins)

Production process:
Large batch production, make to stock
Procurement of ingredients mainly from local suppliers
Quality control (batch) with uncertain outcome, 2-3 days

Distribution:
Company supplies large and medium retailers, drugstore chains, individual chemists
For German market, all products are stored at the production facility
Shipping is outsourced
Due to (increasing) demand variance, supply shortages and uncertain manufacturing yields, company
increasingly experiences stock-outs

52
Case Data

Customer specific contracts and conditions:

When negotiating contracts with large customers (large retailers and large drugstore
chains), company has little bargaining power

Virtually all requirements have to be met, sometimes drastic penalties or consequences


if orders are not fulfilled according to requirements, examples:
Complete shipments are being sent back
penalty = sales volume of order
price discounts enforced in next round of negotiations

Large customers can order whatever and whenever they want and can specify delivery
date

53
Case Data

Order processing:

Orders received through multiple channels (phone, fax, E-mail), now also online portal

Most orders are printed and entered manually in SAP system

Orders processed first come first served, some orders expedited, if very short delivery
lead-time

Availability is being checked in SAP system, customer receives confirmation after order
has been processed

No booking levels, only certain quantity is reserved for one very nasty customer

54
Case Data

One specific stock-out situation:

October/November 2002

Product: Cough drops

Reason for stock-out: supply shortage (quality issues of one supplier) in combination
with high demand due to wave of flu

Upcoming stock-out situation could have been anticipated one week prior, but no active
management of remaining inventory

At 10/23, 150 orders (4071 units) committed with delivery dates

Between 10/23 and 11/11 656 orders (8660 units) received

55
Case Data

1800
1600
Ordered Quantity
Quantity Committed Previously
1400

Units 1200
1000

800
600
400

200
0
23-10

24-10

25-10

28-10

29-10

30-10

31-10

1-11

4-11

5-11

6-11

7-11

8-11

11-11

12-11

13-11

14-11

15-11

18-11

19-11
Dates

56
Inventory and Fulfillment Situation
Companys Results

8000
Planned Receipts Total Quantity Ordered Available Quantity
7000
6000
5000
Units

4000
3000
2000
1000
0
23-10

24-10

25-10

28-10

29-10

30-10

31-10

1-11

4-11

5-11

6-11

7-11

8-11

11-11

12-11

13-11

14-11

15-11

18-11

19-11
-1000
-2000
Dates

57
Rejected Orders Companys Results

Total number of orders Rejected

500

400

300

200

100

0
1 2 3 4 5
Customer Priority

Is this the profile we want?

58
Alternatives

SAPs methodology:

Availability is checked through ATP-function in SAP

For main customers, future orders are forecasted

If forecasted orders exceed available quantity, certain incoming orders are blocked
(rejected?)

Only orders from high priority customers can be processed

What are advantages/disadvantages of this approach?

59
Alternatives

Similar: Booking levels

Reserve portions of available inventory for certain customer classes


Dynamic service level approach how do we determine quantities to be reserved?
What are the prerequisites?
8000
Planned Receipts Total Quantity Ordered Available Quantity
7000
6000
5000
Units

4000
3000
2000
1000
0
23-10

24-10

25-10

28-10

29-10

30-10

31-10

1-11

4-11

5-11

6-11

7-11

8-11

11-11

12-11

13-11

14-11

15-11

18-11

19-11
-1000
-2000
Dates

What are the advantages/disadvantages of a booking level approach?

60
A Different Approach:
Batching and Partial Deliveries

Apply batching and partial deliveries: Utilize customers flexibility in terms of


response time and
delivery quantities

Customers do not require same day order confirmation

Some customers are willing to accept partial deliveries

61
Experiment 1: 1-Day-Batching,
Partial Deliveries Allowed
Unfilled orders in %

70
Company's results Model's results
60
50
40
30
20
10
0
1 2 3 4 5 Overall
Customer Priority

Partial deliveries: 1.22 %

Not very exciting! What are the reasons? What are the
implications?

62
Experiment 2: 1-Day-Batching,
Partial Deliveries Enforced

Partial deliveries are enforced: If customer generally agrees to partial deliveries,


all orders will be fulfilled with two deliveries if possible (see model assumptions
and constraints)

Impact on available inventory:


8000
Planned Receipts Committed and Unfulfilled Quantity Available Quantity Partial Delivery Quantities
7000
6000
5000
Units

4000
3000
2000
1000
0
23-10

24-10

25-10

28-10

29-10

30-10

31-10

1-11

4-11

5-11

6-11

7-11

8-11

11-11

12-11

13-11

14-11

15-11

18-11

19-11
-1000
-2000
Dates

63
Experiment 2: 1-Day-Batching,
Partial Deliveries Enforced

Results: Unfilled orders


Company's results
Results Experiment 1
Results Experiment 2
70
60
50
40
30
20
10
0
1 2 3 4 5 Overall
Custome r Priority

Partial deliveries:
Customer Priority 1 2 3 4 5 Overall
Partial Deliveries in % 0,00 3,96 0,00 9,52 48,28 7,62

64
Further Analysis

Impact of batching: Two and three day intervals


Accepted by customers? Is the achieved benefit worth it?
Problem: Some orders have to be fulfilled next day. How do we handle these?

How strong is the impact of partial deliveries?


Does it make sense to try to negotiate partial deliveries with customers?
Do we want to compensate them for accepting these?

Are there any other ways of creating flexibility in terms of customers


requirements which can be utilized in stock-out situations?

Are there more intelligent ways of allocating the available inventory to


customers?

To which extent do these results apply to other stock-out situations? Can they be
generalized?

65
Readings

Pibernik, R. (2005). Advanced available-to-promise: Classification,


selected methods and requirements for operations and inventory
management. International Journal of Production Economics, 93-94(1),
239-252.
Kilger, C./Meyr, R. (2010). Demand Fulfilment and ATP, in: Stadtler,
H./Kilger, C. (eds.): Supply Chain Management and Advanced Planning,
Berlin & Heidelberg, Chapter 9.

Pibernik, R. (2006). Managing stock-outs effectively with order fulfillment


systems. Journal of Manufacturing Technology Management, 17(6), 721-
736.
(Note: This article provides the details of the case study presented on
slides 52-65; you are not required to study the formal details)

66