Inventory Forecasting

Art?
Science?
Crystal Ball?
or ………..?

Presented by:

George J. Mollo, Jr.

GJM Associates
NCOF, May 2001

 2001, George J. Mollo, Jr.

 Topics to be Covered

• At least 10 factors to be considered

when forecasting
• Several methods of forecasting
• Ways to forecast new items and
repeat items
• How to improve inventory coverage to
achieve maximum fill rates
• Profitability considerations
• Ways to simplify the process

 2001, George J. Mollo, Jr.

 Inventory Forecasting
Art?
Science?
Crystal Ball?

or ………..?

UNDERSTANDING
RELATIONSHIPS !!
 2001, George J. Mollo, Jr.
 How Do We
Improve
“Forecasting” Accuracy?

 2001, George J. Mollo, Jr.

 Simple & General Approach

• Improve majority of items - Don’t try for

100% accuracy on all items
“ALMOST” Right NOW is better than…
“EXACTLY” Right Later !

• Reduce exceptions
• Take Small steps

 2001, George J. Mollo, Jr.

 Planning

• Formal Planning Process

• Historical Analysis
• 6-12+ months in advance
• Assortment Plans (Category, etc.)
• Category/Sub-Cat Relationships

 2001, George J. Mollo, Jr.

 Planning

• Item Counts (based on pages/density)

• New vs. Repeat Item relationships
• Space Allocation
• Margin Targets & Price Points
• Average Price Offered vs. Sold
• Vendor Performance History
• Estimated Item Profitability

 2001, George J. Mollo, Jr.

 Communication (Pre & In Season)

• Marketing (demand by drop, list

information, target audience, etc.)
• Merchandising (market, competition,
styles, etc.)
• Creative (Coordinate presentation to
audience, availability, add/drop items)
• Internal Relationships - Between
Departments

 2001, George J. Mollo, Jr.

 It’s All in RELATIONSHIPS !

• How Many Use These Measures –

Yet Independent of Each Other ??

• Category Share
• Item Count
• Page Space
• Demand (Book & Item)
• Page or Spread Average
• Etc…...

 2001, George J. Mollo, Jr.

 Forecasting Factors

• Presentation (Size) • Item Counts

• Creative Factors (Density)
• Location in Catalog • Space
(key positions) • Space Costs (Sell
• Competition (Internal Ratio)
& External) • Item Relationships
• Demand History (Ranking A - B - C)
• Price Points • Seasonality
• Circulation • Indices
(Qty/Target)

 2001, George J. Mollo, Jr.

 Methods of Forecasting

• Units
• Who’s unit forecast number?
• Dollars
• Forecast to dollars, buy to units
• $ per k circ **
• AII (Average Item Index)

• Forecasts “tied out” – Top Down

AND Bottoms Up
 2001, George J. Mollo, Jr.
 Methods of Forecasting

• Other “Tools” - or “Simple Checks”

• Productivity Index (Category)

• Space Index (Category)
• Ad to Sales Ratio (Sell Ratio)
• Page Average
• Top Item Forecast vs. Bottom Item
Forecast Relationship (within a
dept.)

 2001, George J. Mollo, Jr.

 $ per 1000 Circulated

(Projected item $$ demand /

circulation (k))

• Example: Item A Projected $23,000

demand / circulation 3.5mil)
• $23000 ÷ 3500 = $6.57 per 1000 circ.

** Caveat

 2001, George J. Mollo, Jr.

 $ per 1000 Circulated

• Advantage: Consistent Item Comparison

Base within same drop/catalog

• Disadvantage: Inconsistent Item

Comparison Base between drops/catalogs
• Marketing changes to “productivity” must be considered

 2001, George J. Mollo, Jr.

 AII (Average Item Index)

AII= Projected Item $$ Demand ÷

(Projected $$ demand in Offer/
Number of items)

 2001, George J. Mollo, Jr.

 AII (Average Item Index)

Projected Catalog Demand: $5,000.0

Total # of Items: 250

Average Item: $20,000

Item A = $ 30,000 1.5

Item B = $ 20,000 1.0
Item C = $ 14,000 0.7

 2001, George J. Mollo, Jr.

 AII (Average Item Index)

• Advantage: Consistent Item Comparison

Base within same drop/catalog

• Advantage: Consistent Item Comparison

Base between/across drops/catalogs

• Advantage: When forecast changes –

Index doesn’t change, only value does

• Demonstrates Relationships (To Overall

Average & Other Items)

 2001, George J. Mollo, Jr.

 AII (Average Item Index)

• Disadvantage: Space must be factored

(full page, cover vs. average)
• Disadvantage: Seasonality must be
factored

*However, Space and Seasonality changes

are easy to identify

 2001, George J. Mollo, Jr.

 Average Item Index

Item # Description drop 1 drop 2 drop 3 drop 4 drop 5 drop 6

Item A l/s shirt 2.0 1.8 1.6 1.3

Item B s/s shirt 0.4 1.4 2.0 5.0 2.5 1.8
Item C plaid shirt 1.0 0.8 1.0
Item D henley shirt 0.9 1.0 1.0
Item E pique polo 2.0 2.0 2.2 2.1 2.1 2.0
Item F interlock polo 2.0 1.9 1.8 1.9 1.9 1.9
 2001, George J. Mollo, Jr.
 Average Item Index

Item # Description drop 1 drop 2 drop 3 drop 4 drop 5 drop 6

Item A l/s shirt 2.0 1.8 1.6 1.3

Item B s/s shirt 0.4 1.4 2.0 5.0 2.5 1.8
Item C plaid shirt 1.0 0.8 1.0
Item D henley shirt 0.9 1.0 1.0
Item E pique polo 2.0 2.0 2.2 2.1 2.1 2.0
Item F interlock polo 2.0 1.9 1.8 1.9 1.9 1.9
 2001, George J. Mollo, Jr.
 Other Measures/Indices
(Trends/Profitability Indicators)

• Performance Index (Category AII)

• Category Space Index
• Sell Ratio

 2001, George J. Mollo, Jr.

 Performance Index:

(% of demand / % of items)

(a.k.a. - Category AII)

 2001, George J. Mollo, Jr.

 Performance Index

This Year This Year TY TY

Description Items % Demand % Avg. Perf.

Spring 99
TOTAL DROP: 218 100.0% 5500.0 100.0% 25,229 1.00

GIFTS 43 19.7% 1204.5 21.9% 28,012 1.11

HOME 54 24.8% 1353.0 24.6% 25,056 0.99

APPAREL 50 22.9% 1743.5 31.7% 34,870 1.38

ACCESSORIES 71 32.6% 1199.0 21.8% 16,887 0.67

 2001, George J. Mollo, Jr.

 Category Space Index:

% of demand / % of space

 2001, George J. Mollo, Jr.

 Space Index
This Year This Year TY
Description Pages % Demand % Perf.

Spring 99
TOTAL DROP: 64 100.0% 5500.0 100.0% 1.00

GIFTS 20 31.3% 1204.5 21.9% 0.70

HOME 14 21.9% 1353.0 24.6% 1.12

APPAREL 16 25.0% 1743.5 31.7% 1.27

ACCESSORIES 14 21.9% 1199.0 21.8% 1.00

 2001, George J. Mollo, Jr.

 Sell Ratio:
(Item percent of page * cost per
page) / Projected item $$ demand
in Offer

 2001, George J. Mollo, Jr.

 Sell Ratio

Projected Catalog Demand: $5,000.0

Total Pages: 64
Total Selling Costs: $1,370.0
(Creative, Paper, Printing, Postage, List Rental)
Selling Cost Per Page: (Editorial ?) $21,406

Item A = 25% of Page ($30k Demand) 17.7%

Item B = 25% of Page ($20k Demand) 26.5% Sell Ratios
Item C = 25% of Page ($14k Demand) 38.0%
(25% page = $5351.5 Ad Cost)
 2001, George J. Mollo, Jr.
 Department View

Average item 25,229

Page Cost: 11,983

Description Cost Retail MU Dmd AII % Sell

% $$ space Ratio

l/s Shirt 7.50 18.00 58.3% 15.7 0.62 0.25 19.1%

s/s shirt 5.50 12.00 54.2% 22.7 0.90 0.25 13.2%
plaid shirt 7.80 20.00 61.0% 11.4 0.45 0.25 26.3%
henley shirt 8.00 22.00 63.6% 37.0 1.47 0.50 16.2%
pique polo 10.00 32.00 68.8% 32.2 1.28 0.40 14.9%
interlock polo 10.00 32.00 68.8% 25.3 1.00 0.25 11.8%
denim shirt 9.15 27.00 66.1% 18.5 0.73 0.25 16.2%
chambray shirt 15.00 38.00 60.5% 16.7 0.66 0.25 17.9%
dress shirt 17.75 40.00 55.6% 20.2 0.80 0.25 14.8%

Department Totals: 199.7 2.65 15.9%

Avg Price Offrd 26.78
Avg Price Sold 23.65 62.6%

 2001, George J. Mollo, Jr.

 Forecasting Repeat Items

• What % of items will be planned

for Repeat vs. New Items
• Review History
• Identify Similar Items
• Consider “Wear Out” factors
• Review Relationships to other product
(especially competing new product)
• Profit Projections (Is the item still justified?)
• Suggestion: Add inventory carrying cost to
margin calculation…

 2001, George J. Mollo, Jr.

 Forecasting New Items

• Identify Similar Items

• Varying Rules -- New Items should be
> 20% of average
• Review Relationships to other product
• Consider competition (Internal & External)
• Profit Projections (Is the item justified?)

 2001, George J. Mollo, Jr.

 Other Forecasting Factors

• Final Item Forecasts should NEVER be

done in isolation
• Sum of the parts will ALWAYS be Greater
than the Whole = OVERSTOCK
• When using “trend curves” generally
3 to 5 will suffice (Keep it simple!)
• Forecasts should be done by drop AND
revise balance of season (use of AII
makes this relatively easy)

 2001, George J. Mollo, Jr.

 Average Item Index

Item # Description drop 1 drop 2 drop 3 drop 4 drop 5 drop 6

Item A l/s shirt 2.0 1.8 1.6 1.3

Item B s/s shirt 0.4 1.4 2.0 5.0 2.5 1.8
Item C plaid shirt 1.0 0.8 1.0
Item D henley shirt 0.9 1.0 1.0
Item E pique polo 2.0 2.0 2.2 2.1 2.1 2.0
Item F interlock polo 2.0 1.9 1.8 1.9 1.9 1.9
 2001, George J. Mollo, Jr.
 Other Forecasting Factors

• Life of Item (> 1 Campaign)

• Realistic Estimating **
• Based on Demand NOT on Availability
• Maintain Relationship to Buy Plan
• Don’t Spend $$ Twice
• Assortment Plan is the CONTROL

 2001, George J. Mollo, Jr.

 Improving Inventory Coverage

• Better Planning & Forecasting allows

increased “Just in Time” Buying
• Longer Range Planning - Identifies
“Horizon” of Need
• Plan by “Horizon/Season” rather than “Drop”
• Inherent “Creative Merchant” hurdle –
overcome with 80/20 rule
• Planning Reduces “hidden costs”
• Impact on Creative Schedules
• Some DC issues

 2001, George J. Mollo, Jr.

 Improving Inventory Coverage

• Buys can be Scheduled

• Based on “Horizons” and “Lead Times” –
allows greater flexibility for change
• Allows forecasts to be modified to improve
fill rates (lower backorder costs, lower
overstock, increased turns)

• Risk/Rewards of Overbuying Can be

Considered - In Advance

 2001, George J. Mollo, Jr.

 Internet Forecasting

• New Medium/Channel
• Standard Catalog Benchmarks?
• Demand, Response Rates?

• How to Measure?
• Hits, Click-thru, Traffic, Actual Sales?

 2001, George J. Mollo, Jr.

 Internet Forecasting

• Plan Assortment
• Plan Sell-Thru
• Constantly Moving the Floor – and
Adapting to Sales (Pull items on/off the
Web Page)
• Approach
• Combination of Retail & Catalog – (ratio of
days/views offered to sell-thru/sales at base
level??)

 2001, George J. Mollo, Jr.

 Possible Metrics for eCommerce

• Modify AII to account for Average Days

On Line
• Same relationship as AII, HOWEVER factoring
Days on Line to obtain a NEW Average or
Index

• Modify AII to account for Average Views

On Line

 2001, George J. Mollo, Jr.

 SUMMARY
• Simplify the Process
• Reduce exceptions
• Use Averages
• Use of Indices
• Indices Offer a Simplified Approach
• Indices create a Reference Base of 1.0

• “Simple Checks” (Compare to Various

Averages, Simple references, etc.)

 2001, George J. Mollo, Jr.

 Relationship Reviews !!

Multiple Measures (or Indices) Used

Together - Will Improve the Reliability of
the Forecast & Increase Profitability

For example:
• Performance Index with Space Index
• Space & Item Demand with Sell Ratio
• Average Item with Item Count (and
Demand)

 2001, George J. Mollo, Jr.

 Benefits to Other Areas
(Positive Profit Impacts)

• Inventory Buying can be “Scheduled” on a

“Horizon”
• Horizon forecasts help DC in “bin profiling”
• Dramatic positive impacts to cash flow
• Reduced inbound freight expense (less
expedited shipments)
• Reduced backorders & reduced overstock
• Higher fill rates (both initial and final)
• Increased “Bottom Line”

 2001, George J. Mollo, Jr.

 Inventory Forecasting
Art?
Science?
Crystal Ball?
or ………..?

RELATIONSHIPS !!!!

With Work, Planning Discipline &

Internal Communications
– Forecast Accuracy will Improve !

 2001, George J. Mollo, Jr.

 Inventory Forecasting

Art?
Science?
Crystal Ball?
or ………..
RELATIONSHIPS !!!
Presented by:

George J. Mollo, Jr.

GJM Associates
NCOF, May 2001

 2001, George J. Mollo, Jr.