I2 U Intelligent Supply Chain Management Course Module Six: Demand Planning.

i2 UIntelligent Supply Chain Management

Course

Module Six:Module Six:DemandDemand Planning Planning

2©2001 i2 Technologies Inc CONFIDENTIAL

Supply Chain Management Key Processes

Sales & Operations Planning

Demand Planning

Inventory Planning

Supplier Scheduling

Production Scheduling

Inventory Deployment

Transportation Scheduling

Demand Fulfillment

Supply Chain Execution Monitoring

Strategic Supply Chain

Planning

Master Supply Planning

Procurement

Production Distribution

Transportation

Impact of decisions

_

+

Num

ber o

f dec

isio

ns

+

_

Spec

ifici

ties

by in

dust

ries Length of Planning horizon

Fu

lly

Inte

gra

ted

to

p-d

ow

n d

irec

tio

ns

Fu

lly

Inte

gra

ted

bo

tto

m-u

p f

eed

bac

k

Reaction to changing supply conditions


After Completing This Module,You Are Expected to:

Define the key business objectives of the Demand Planning process

Apply “focused forecasting” and “consensus forecasting” logics

Understand the multi-dimensional nature of the Demand Planning process and the underlying enabling technology available

Identify Demand Planning key enablers and their resulting business value

Identify Demand Planning excellence criteria


hou

rs

days

wee

ks

mon

ths

year

+

buy

make

move

sell

store

operational tactical strategicscheduling

Demand Planning Process Positioning


t

HISTORY FORECASTS

Promotion

Price increaseDownsizing of market expectations

Stock-out

Promotion End of quarter

‘Sales Forecast are Always Inaccurate


Clean up historical data Selection of the right forecasting technique Capture all intelligence influencing future demand

– promotions / pricing modifications– phase-in/phase-out

Compare inputs coming from different players using different languages

– Sales manager: sales by region– Marketing manager: sales by product family– Sales rep: sales by clients– Finance/management: global sales– statistical forecast: sales by SKU

Agree on a consensus forecast

Major Challenges of the Demand Planning Process


Selection of the Right Forecasting Technique Through the ‘Focused Forecasting’ Logic

time

HISTORY FORECASTS

today- 3 years

-1 year

… each technique is used to generate a “forecast” over the

recent past 2

To identify which forecasting

technique is the right one...

1

The technique with the least error margin with actual sales is used to forecast future sales

3


Clean up of historical data Selection of the right forecasting technique Capture all intelligence influencing future demand

– promotions / pricing modifications– phase-in/phase-out

Compare inputs coming from different players using different languages

– Sales manager: sales by region– Marketing manager: sales by product family– Sales rep: sales by clients– Finance/management: global sales– Statistical forecast: sales by SKU

Agree on a consensus forecast

Major Challenges of the Demand Planning Process


The Challenge of Capturing, Comparing and Reaching Consensus on a Unique Sales Forecast

Last Year’s Q3was flat because

the weather was wet

The new Wonder Gizmo will hit

the market in June

We will increase salesof big ones by 10%

in Q3 and Q4

My shop will sell 10additional Widgets and

7 Gizmos in JanuaryIf we sell through

e-commerce we can sell 25,000 more but will reduce direct sales by 10%

Any new product cannibalizes demand

for the old one

Our Mega competitormay launch his new

Widget in May


Forecasting is a Multi-Dimensional Problem


Describing a Product is Multi-Dimensional

All Products -Beverages

Subgroup -Cola -Ginger Ale -Root Beer

Group - Soft Drinks - Ice Teas - Juices

Product -Cola 16 oz. -Diet Cola 16oz. -Cola 32 oz. -Diet Cola 32 oz.

Packaging -Glass -Can

Size -16oz. -32oz. -48oz.

Style -Diet -Regular


Describing Time is Multi-Dimensional

WeeklyWeekly

MonthlyMonthly

QuarterlyQuarterly


Demand Planning Reality

Various people working in various corporate and product dimensions, in various time increments, using different data and creating different plans

These people need to collaborate with one another

Such collaboration requires a multi-dimensional environment


“Top-down”

Budget/Long Range Sales Plan

Key Account Plans

“Bottom-up”

Management• Forecast

Consolidation• Forecast Generation• Adjustments

Geography

Time

Product Product Management/ R&D•Ramp-ups•Ramps-down•Product Life Cycle

Sales Management• Growth/Trends• Sales Actions/

Promotions• New Product Intro

Data Consolidation•Customer Orders•Shipments

Demand Planning Collaboration Requires a Multi-Dimensional Database


Product dimension

Geography dimension

Several forecasts or related

information


Monthly or Weekly CycleMonthly or Weekly Cycle

Market Market IntelligenceIntelligence

Demand Demand HistoryHistory

Statistically Statistically Derived Derived ForecastForecast

Consensus Forecast MtgConsensus Forecast Mtg

SalesSales

MarketingMarketingFinanceFinance

ManufacturingManufacturing

Typical Collaborative Forecasting Process


Example of Multi-Dimensional Complexity:IBM PSG Demand Planning Project

Worldwide Implementation Covering US, Asia Pacific, Canada, Latin America, Europe, Middle East and Africa

48 Users 8,000 SKUs 40 Levels 11 Views 163 Rows


Collaborative Demand Planning With Customers

Shared applica

tion

Retailer

Manufacturer

Comparison of both inputs to reach

consensus on sales figures3

Manufacturer publishes its sales forecast, which

includes trade promotions and

advertising campaigns

1

Retailer publishes its forecast requirements, which include

sales promotions and merchandising actions (shelf

space management, delisting, …)

2


Demand Planning Process Key Enablers and Related Business Benefits

Best of breed statistical forecasts + “pick best” capability

All influences on future demand captured in a multi-dimensional database

Consensus forecasting enabled by the comparison of multiple forecasts

Collaborative planning tool enabling to compare demand with customer plans

Forecast accuracy

Customer service

Inventory turns


Major Roadblocks to Remove in Order to Implement a Best-in-Class Demand Planning Process

Reliability of historical data– history cleaned from past promotions, atypical orders and

stockouts

Availability of historical data at the required level of granularity

Lack of accountability and incentives on forecast accuracy goal– sales people tend to overestimate potential sales to ensure

maximum availability

– or these same people tend to underestimate potential sales to ensure that sales objectives will be met


Demand Planning Process Excellence Criteria (1/2)

The frequency of the demand planning process is adapted to the level of demand variability.

A structured workflow organizes the activities of the demand planning process Past data used to extrapolate future demand is reliable and cleaned from

atypical sales conditions (stockouts, promotions…). Times series are long enough to generate reliable extrapolations and trends.

The application of the ‘focused forecasting’ method ensures that the right statistical forecasting technique is applied for each product and each level of aggregation. The people that manipulate these statistical forecasting techniques have a strong expertise in this domain.

As many ‘demand pyramids’ as necessary (with as many aggregation levels as necessary) are modeled in a multi-dimensional database supporting the demand planning process, to enable all people involved in demand planning to adequately provide their input on future demand.

All people with visibility on demand future evolution and/or with activities impacting future sales systematically contribute to the demand planning process.


The impact of promotions, shelf management, category management and advertising activities is taken into account in the forecast.

Supply restrictions and allocations by sales channels are taken into account in the final sales plan.

Whenever customers plan their purchases, this information is received directly from them. It either eliminates the need to forecast, or it is compared to the forecast generated internally (in this case, when differences show up, a final number is agreed to by both parties). The company is actively pursuing the expansion of such collaborative demand planning processes.

Forecast of future products are validated jointly by Marketing and R&D. R&D has adequate planning processes that ensure the reliability of the new product launch schedules.

Product phase-out decisions take into account possible costs attributable to unused inventories.

All forecasts are reconciled, and a unique, final consensus demand plan is generated, which is used to drive ALL Supply Chain Management activities.

Demand Planning Process Excellence Criteria (2/2)

I2 U Intelligent Supply Chain Management Course Module Six: Demand Planning.

Documents

inaccurate2001 i2 technologies

years2001 i2 technologies

may2001 i2 technologies

demand planning processclean

direct sales

regionmarketing manager

different players

new product