Learn more about QAD`s new acquisition

Report
QAD West Coast User Group
Demand Planning 7.0
Mark K Williams, CFPIM, CSCP
February 12, 2013
QAD Global Supply Chain Planning Practice
Introduction
Agenda
• Introduction to DynaSys
• Introduction to forecasting
• Demand Planning 7.0 attributes
• Easy On Boarding
2
Introduction to DynaSys
3
Introduction to DynaSys
The Mission of DynaSys
“Offer to our customers,
a Comprehensive and Effective Planning Solution
dedicated to
Supply Chain Excellence,
benefiting from our 27 years of expertise.”
Durability
Expertise
Partnership
Loyalty
A Few DynaSys Customers
5
Introduction to DynaSys
Global Customers With Multiple Sites
Ipsen : 110 sites
Essilor : 18 sites
Everris : 32 sites
Mapa/Spontex : 27 sites
Baxter Healthcare : 25 sites
Introduction to DynaSys
DynaSys Customer Base
Luxury
14%
Industry
14%
Retail
19%
Food &
Beverage
39%
Pharma
14%
7
Introduction to DynaSys
Master Planning
Network & Inventory Optimisation
S
U
P
P
Procurement
L
Planning
I
E
R
Not Yet!
Production
Planning
Distribution
Planning
Demand
Planning
C
U
S
T
O
M
E
R
Operational
Tactical
Strategic
Supply Chain Planning Suite
Procurement
Production
Distribution
Sales
70
60
50
Upper Limit
Forecast
Lower Limit
40
30
20
10
0
1st Qtr
2nd Qtr
3rd Qtr
4th Qtr
Introduction to Forecasting
Importations
Corrections
Calculation
history
Life cycle
Metrics &
validation
collaboration
Forecasts
Detailed monthly forecasting process
9
Introduction
An Industry Supply Chain
Physical Supply Chain
Natural
Resources
Suppliers
Production
Distribution
Retail
Information Supply Chain
Consumers
Introduction
What is a Forecast?
An estimate of future demand. A forecast can be
constructed using quantitative methods, qualitative
methods, or a combination of methods, and it can be
based on extrinsic (external) or intrinsic (internal)
factors. Various forecasting techniques attempt to
predict one or more of the four components of
demand: cyclical, random, seasonal, and trend.
APICS Dictionary
4 Components of Demand
Seasonal
Trend
Daily (fast food)
Annually (snow plows)
New Product
New!
Follows Economy
Mature
Product
Outlier
Random
Cyclical
Introduction
Forecasts Are More Accurate When Aggregated
How Many Brown,
Size 10 Shoes
Will We Sell
Thursday?
Not Very Accurate –
Just a Wild Guess
How Many
Shoes Will We
Sell This
Month?
Better Chance Of Being
Accurate – Can Use Data
Introduction
Forecasts Become Less Accurate With Time
70
60
50
Upper Limit
Forecast
Lower Limit
40
30
20
10
0
1st Qtr
2nd Qtr
3rd Qtr
4th Qtr
Introduction
To Improve Forecast Accuracy, Involve Your
Customers & Major Stakeholders
It All Sounds Good, But….
16
Introduction
Sr. Managers Meet To Discuss Forecasting
17
Introduction
Senior Management Discussion Overheard…
• “We have 42,000 SKU’s and a limited budget for
personnel to do forecasting. How do we get the
most bang for the buck??”
• “Sales, Marketing & Demand Planning all think they
are better at forecasting promotions. Who is right?”
• “My old company kept 2 weeks of Safety Stock on
hand for each item. Isn’t that enough?”
• “Which products are behind budget?”
• “The forecast is always wrong, why bother?”
18
Senior Management Overheard….
“The Forecast Is Always Wrong, Why Bother?”
19
Why Forecast?
Forecasting Rule #1
The Forecast Is WRONG
Why Forecast?
Why Forecast?
• To anticipate change
-
Promotions
New products
Product retirements
Customer buying habits
• To anticipate inventory and capacity demands
• To procure materials in a timely manner
• To project costs of operations into budgeting
processes & manage cash flow
Why Forecast?
The Quandary
Supplier Lead Time
East Coast
DC
Overseas
Supplier
Customer Lead Time
Item
Scanned
Delivery
Expected
22
Why Forecast?
Companies That Do Not Adequately Forecast:
• Often have employees that work overtime for
months, followed by layoffs
• Are poor users of manufacturing capacity
• Are unable to fulfill “surprise” customer orders
• Face huge expediting charges
• Have excessive inventory write-offs
• Have the dreaded combination of a poor customer
service level and too much inventory!
• Disappoint customers by not responding to
promotions on time
Introduction
Out of Stock Rates Double on Promoted Items
Average On-Shelf Availability Levels
Frozen Pizza
75.3%
90.1%
89.5%
Soft Drinks
86.1%
Salty Snacks
85.8%
Cookies & Crackers
95.0%
88.7%
Milk
85.1%
Total
70%
Source: Stax, Inc. Store Audits
94.0%
94.0%
79.4%
Beer
93.9%
75%
80%
85%
Promoted Items
96.8%
92.7%
90%
All Items
95%
100%
Why Forecast?
Value Proposition and ROI
• Reduced costs
- Inventory reductions 8-10% in 6 months, 12-70% over 2-3 yrs
- Reduced obsolescence and inventory write-offs 30-50%
- Total cost reductions 1-2%
• Improved Performance
-
On-time delivery to customers improvement 10-40%
Plant efficiency improved between 2 and 33%
Improvement in forecast accuracy – 5-20%
Improved profitability 1-3%
Source: IIF and other industry sources
25
Senior Management Overheard….
“We have 42,000 SKU’s and a limited budget for
personnel in forecasting. How do we get the most
bang for the buck??”
26
Overcoming Limited Resources
DP 7.0 Answers….
•
•
•
•
•
•
Isolating importation errors
Identification of items behaving badly
Identification and correction of historical anomalies
Superior Forecast Engine
ABC Analysis to separate the critical from the trivial
Multi-level forecasting
27
QAD Demand Planning 7.0
Statistical Forecasting Interface
C
o
n
t
r
o
l
s
Graphs
Excel
Like
Grid
Collaborative
Messages
S
h
o
r
t
c
u
t
s
Overcoming Limited Resources
Isolating Importation Errors
29
Introduction
Importation & Correction
•
•
•
When importing data,
problems can occur
DP 7.0 makes it easy to
identify problems that
the user can fix
DP 7.0 automatically
highlights items with high
forecast error – major
time saver
30
Isolating Importation Errors
Identifying Import Problems
1
3
1. Click items without
standard cost
2. Select Unit
Conversion Issues
3. The list of 30 items
appears
Fix a problem once, &
it is fixed permanently
2
31
Isolating Importation Errors
Identifying Import Problems
1
2
3
32
Overcoming Limited Resources
Identification Of Items Behaving Badly
33
Identification of Items Behaving Badly
Last Month’s Alerts Highlight Forecast Problems
Last Month’s
Forecast Was Very
Wrong – User
Should Review
34
Identification of Items Behaving Badly
Modifying Alert Parameters
• Deviation Gap set at 10% will generate no alert
• Change Deviation Gap to 5% generates an alert
35
Overcoming Limited Resources
Identification And Correction Of Historical
Anomalies
36
Identification & Correction of Historical Anomalies
Why Modify History?
• Forecasts are built on pattern recognition
• What if your sales history included:
- Extraordinary demand caused by a competitor’s
stockouts?
- A weather driven surge or absence of demand?
- A drop in orders caused by your prolonged stockout?
- Would you want the future to reflect that past?
• DP 7.0 provides statistical boundaries that help
identify these anomalies, allowing you to accept or
modify them
37
Identification & Correction of Historical Anomalies
Outliers In Sales History
38
Identification & Correction of Historical Anomalies
Options for Outliers
A. Item looks
suspicious
B. Modify
history
C. Accept
history
Actual History
Modified History
Suspicious Quantity
Modification
Validation
39
Overcoming Limited Resources
Superior Forecast Engine
40
Superior Forecast Engine
Forecast Pro
• Method used to select the “best fit” time series
forecasting techniques for your data
• Judged to be one of the most accurate in the
world
• There are over 30 other forecasting methods
available including:
-
Curve Fitting
Exponential Smoothing
Regression
New Product
Lifecycle
• One size DOES NOT fit all!
Overcoming Limited Resources
ABC Analysis To Separate The Critical From The
Trivial
42
ABC Analysis To Separate The Critical From The Trivial
ABC Inventory Management
• First noted by Vilfredo Pareto
• Items are classified as:
- A items (fastest movers)
• 20% = 80% of sales or usage
- B items (middle-range movers)
• 30% = 15% of sales or usage
- C items (slow movers)
• 50% = 5% of sales
• Focus on fast moving items
• Useful in many applications
- In forecasting, where we want to keep our attention
ABC Analysis To Separate The Critical From The Trivial
Tighter Alert Parameters For “A” Items
• Deviation Gap set at 10% will generate no alert
• Change Deviation Gap to 5% generates an alert
44
ABC Analysis To Separate The Critical From The Trivial
ABC Classification
A=80%
9
I
t
e
m
s
ABC Analysis To Separate The Critical From The Trivial
ABC Classification
A=60%
6
I
t
e
m
s
Overcoming Limited Resources
Multi-Level Forecasting
47
Multi-Level Forecasting
Why Not Forecast At The Item Level?
•
•
•
•
Number of SKU’s
Aggregation improves accuracy
Forecasting resources
Collaboration
48
Multi-Level Forecasting
Bottom Up – Top-Down Forecasting
Product
Group
49
Multi-Level Forecasting
Determine Difference Lower & Upper Forecasts
Item
XR-24
XR-34
::
::
::
Total XR PG
Total BH PG
Total TW PG
Total Business
Unit 1
Total Business
Unit 2
Total Company
Forecast
(Units)
2,876
4,578
91,684
276,908
39,489
Corporate Sales Forecast
Percent Difference
Cost
$ 2.98
$ 4.98
$
$
Dollar
Forecast
8,570.48
22,798.44
$
$
$
1,127,715.66
3,465,098.00
1,975,843.00
$
6,568,657
$
$
3,972,805
10,541,462
$
12,000,000
13.84%
Multi-Level Forecasting
Revised Forecast
Item
XR-24
XR-34
::
::
::
Total XR PG
Total BH PG
Total TW PG
Total Business
Unit 1
Total Business
Unit 2
Total Company
Forecast
(Units)
2,876
4,578
Original
Forecast
$ 8,570.48
$ 22,798.44
Revised
Forecast
$ 9,756.31
$ 25,952.88
91,684
276,908
39,489
$ 1,127,716
$ 3,465,098
$ 1,975,843
$ 1,283,749
$ 3,944,536
$ 2,249,225
$ 6,568,657
$ 7,477,510
$ 3,972,805
$ 10,541,462
$ 4,522,490
$ 12,000,000
Senior Management Overheard….
“Sales, Marketing & Demand Planning all think they
are better at forecasting promotions. Who is right?”
52
Forecasting Promotions
Collaboration in the Demand Planning Process
Marketing
Input
Customer
Input
Sales Input
Statistical
Forecast
Demand
Plan
New Product
Development
Forecasting Promotions
DP 7.0 Answers
• Enable Collaboration
• Retain each collaborators input
• Measure each collaborators accuracy
54
Forecasting Promotions
Original, Non-Promotional Forecast
Original Forecast
27,271
55
Forecasting Promotions
1st Collaborator – Marketing
Marketing Forecast
35,000
56
Forecasting Promotions
2nd Collaborator - Sales
Sales Forecast
60,000
57
Forecasting Promotions
3rd Collaborator – Demand Planning
DP Forecast
45,000
58
Forecasting Promotions
When Forecasting Promotions
• Each collaborator inputs their forecast
• A predefined hierarchy determines which forecast
has precedence in determining the final forecast
• All forecasts are stored in the system
• After the promotion the system can determine
which of the forecasts was the most accurate
59
Senior Management Overheard….
“My Old Company Kept 2 Weeks Of Safety Stock On
Hand For Each Item. Isn’t That Enough?”
60
Calculating Safety Stock
How Do You Determine Safety Stock Levels?
Production Qty = 1 Month Average
Safety Stock = 2 Weeks
Part No.
Forecast
Actual
Jan Feb Mar Apr May Jun Total
98 96 103 99 105 99
600
98 96 103 99 105 99
600
Average
100
100
• Based on producing 1 months average sales and
the “2 Week Safety Stock” philosophy, is the
amount of inventory appropriate? Could it be
improved?
61
Calculating
How Do You Determine Safety Stock Levels?
Production Qty = 1 Month Average
Safety Stock = 2 Weeks
Part No.
Forecast
Actual
Jan Feb Mar Apr May Jun Total
98 96 103 99 105 99
600
285 40 220 30 10 15
600
Average
100
100
• Like the previous example, the forecast & actual
totaled 600 for six months
• Based on producing 1 months average sales and
the “2 Week Safety Stock” philosophy, is the
amount of inventory appropriate? Could it be
improved?
62
Calculating Safety Stock
How Do You Determine Safety Stock Levels?
Part No.
Forecast
Actual
Jan Feb Mar Apr May Jun Total
98 96 103 99 105 99
600
98 96 103 99 105 99
600
Average
100
100
Part No.
Forecast
Actual
Jan Feb Mar Apr May Jun Total
98 96 103 99 105 99
600
285 40 220 30 10 15
600
Average
100
100
• The best way to calculate safety stock is based on:
- Forecast accuracy
- Desired service level
- Lead time
63
Calculating Safety Stock
Safety Stock Screen
64
Calculating Safety Stock
Impact of Desired Service Level
65
Calculating Safety Stock
Impact of Differing Lead Times
66
Calculating Safety Stock
Replace Safety Stock In EA
Calculated
Safety Stock
Current
Safety Stock
In EA
• Current Safety Stock imported from EA
• EA Safety Stock compared to Calculated Safety
Stock
• After reviewing the computed Safety Stock levels,
you can have them exported back to EA, or you
can ignore them
67
70
60
50
Upper Limit
Forecast
Lower Limit
40
30
20
10
0
1st Qtr
2nd Qtr
3rd Qtr
4th Qtr
DP 7.0
Easy On Boarding
Importations
Corrections
Calculation
history
Life cycle
Metrics &
validation
collaboration
Forecasts
Detailed monthly forecasting process
68
Easy On Boarding
Why EOB?
DP 7.0 Configurability
Strength
• Can adapt to almost
every business process
• Can be fine tuned to suit
user requirements
DP 7.0
Configurability
Weakness
• Complex
• Longer setup time
• Higher learning
curve for customer
69
Easy On Boarding
What is EOB?
• DP 7.0 is a highly configurable tool
- EOB is designed to meet 90% of customer
requirements “out of the box”
- Reduces implementation time
- Speeds the process of the customer getting value
from the software
• EOB is a preconfigured database that includes:
-
Hierarchies mapped to QAD EA
Standard import / export with QAD EA
Standardized process & parameters
Standardized screens
Standardized KPI’s
70
Easy On Boarding
Integration
Direct ODBC connection
Samba
shared folder
or
FTP transfer
dmtrfcst.csv
QADDP
DataHub
DP 7.0 Import Task
SSIS Package
QAD DP-EA
36.5.9.2
Demand Planning Import
QAD Enterprise
DP 7.0 Export Task
QAD DP 7.0
database
71
Building the
Effective Enterprise
Join us in San Antonio, TX
May 6-9, 2013
Early Bird Ends Soon!
72
Demand Planning 7.0
Contact
Mark K. Williams, CFPIM, CSCP
[email protected]
QAD Global Supply Chain Practice
www.qad.com
© 2013 QAD Inc
73

similar documents