application of the data-cube forecasting to collaborative
TRANSCRIPT
Application of the Data-cube Forecasting to Collaborative Planning and Vendor
Managed Inventory
Igor Gusakov, Ph.D.Moscow, Russia
Easy data access
Business logic
User access rights
Data visualization
Workflow
Data pre-processing
Forecasting
Post-processing (manual adjustments)
Versioning
Reporting/Calculations
Web access/Excel add-in
"Ideal" FSS Typical OLAP
FSS VS. OLAP
Writeback is technology that allows users to publish data back to cubes
WRITEBACK
Writeback is a fantastic tool for judgmental adjustments
Clients Plan
Client_1 10
Client_2 10
Client_3 10
Client_4 10
Client_5 10
Client_6 20
Client_7 20
Client_8 5
Client_9 4
Client_10 1
Summ 100
Clients New plan
Client_1 11
Client_2 11
Client_3 11
Client_4 11
Client_5 11
Client_6 22
Client_7 22
Client_8 5.5
Client_9 4.4
Client_10 1.1
Summ 11010%
Volume Building Blocks (VBBs)
Manual adjustments to forecast made by relevant managers and followed by obligatory comments
Easy way to track who made changes to forecast, and when and why they were made
Avoiding the widespread but counterproductive practice of changing the forecast simply because “the boss said so”
Forecast Value Added methodology to understand who is making the forecast better and who is not
VBBs are company activities or external events that lead to changes in sales volumes
WRITEBACK
WORKFLOW
Cycle – period of time to create a plan
Phase – shorter period inside cycle, when particular users might input particular VBBs
Cycle June 2015
Phase 1
Phase 2
Phase 3
Phase n
Workflow is yet another dimension of the cube
Easy data access
Business logic
User access rights
Data visualization
Workflow
Data pre-processing
Forecasting
Post-processing (manual adjustments)
Versioning
Reporting/Calculations
Web access/Excel add-in
"Ideal" FSS
FSS VS. OLAP
Typical OLAP
Product 1
Product 2
Product 3
Brand 1
Product 4
Product 5
Product 6
Brand 2
All products
Client 1
10
20
30
60
40
50
60
150
210
Client 2
15
25
35
75
45
55
65
165
240
Retail
25
45
65
135
85
105
125
315
450
Client 3
100
200
300
600
400
500
600
1500
2 100
Client 4
150
250
350
750
450
550
650
1650
2 400
Wholesale
250
450
650
1350
850
1050
1250
3150
4 500
All clients
275
495
715
1485
935
1155
1375
3465
4 950
2-D SLICE OF A CUBE
All
Brand 1
Brand 2
Product 1
Product 2
Product 3
Product 4
Product 5
Product 6
Retail
Wholesale
Retail
Wholesale
Retail
Wholesale
Retail
Wholesale
Retail
Wholesale
Retail
Wholesale
Client 1Client 2Client 3Client 4Client 1Client 2Client 3Client 4Client 1Client 2Client 3Client 4Client 1Client 2Client 3Client 4Client 1Client 2Client 3Client 4Client 1Client 2Client 3Client 4
1015
1001502025
2002503035
3003504045
4004505055
5005506065
600650
FORECASTING BY AGGREGATION
VENDOR MANAGED INVENTORY
Primarysales
Secondary sales forecast
Current distributor’s stock
Optimalstock= - +
DistributorManufacturer POS
Primary sales Secondary sales
Collaborative planning is a business model that allows us to combine the information about future activities
of a manufacturer and its distributors
COLLABORATIVE PLANNING
OLAPSecondary sales
forecast
ManufacturersVBBs
DistributorsVBBs
CONCLUSIONS
Although designed for reporting, OLAP is also an effective solution for planning.
VBBs are easily implemented in OLAP tools with Writeback functionality.
OLAP systems with web access are perfect for collaborative planning.
Time-series forecasting required for FSS but absent in a typical OLAP can be effectively implemented with data-cube forecasting.