Inventory Management, Supply Contracts and Risk Pooling

Phil Kaminsky February 1, 2007 kaminsky@[Link]

Issues
Inventory Management The Effect of Demand Uncertainty
(s,S) Policy Periodic Review Policy Supply Contracts Risk Pooling

Centralized vs. Decentralized Systems Practical Issues in Inventory Management

Sources: plants vendors ports

Regional Warehouses: stocking points

Field Warehouses: stocking points

Customers, demand centers sinks

Supply

Inventory & warehousing costs Production/ purchase costs Transportation costs Inventory & warehousing costs Transportation costs

Inventory
Where do we hold inventory?
Suppliers and manufacturers warehouses and distribution centers retailers

Types of Inventory
WIP raw materials finished goods

Why do we hold inventory?

Economies of scale Uncertainty in supply and demand Lead Time, Capacity limitations

Goals: Reduce Cost, Improve Service

By effectively managing inventory:
Xerox eliminated $700 million inventory from its supply chain Wal-Mart became the largest retail company utilizing efficient inventory management GM has reduced parts inventory and transportation costs by 26% annually

Goals: Reduce Cost, Improve Service

By not managing inventory successfully
In 1994, IBM continues to struggle with shortages in their ThinkPad line (WSJ, Oct 7, 1994) In 1993, Liz Claiborne said its unexpected earning decline is the consequence of higher than anticipated excess inventory (WSJ, July 15, 1993) In 1993, Dell Computers predicts a loss; Stock plunges. Dell acknowledged that the company was sharply off in its forecast of demand, resulting in inventory write downs (WSJ, August 1993)

Understanding Inventory
The inventory policy is affected by:
Demand Characteristics Lead Time Number of Products Objectives
Service level Minimize costs

Cost Structure

Cost Structure
Order costs
Fixed Variable

Holding Costs
Insurance Maintenance and Handling Taxes Opportunity Costs Obsolescence

EOQ: A Simple Model*

Book Store Mug Sales
Demand is constant, at 20 units a week Fixed order cost of $12.00, no lead time Holding cost of 25% of inventory value annually Mugs cost $1.00, sell for $5.00

Question
How many, when to order?

EOQ: A View of Inventory*

Note: No Stockouts Order when no inventory Order Size determines policy Inventory
Order Size Avg. Inven Time

EOQ: Calculating Total Cost*

Purchase Cost Constant Holding Cost: (Avg. Inven) * (Holding Cost) Ordering (Setup Cost): Number of Orders * Order Cost Goal: Find the Order Quantity that Minimizes These Costs:

EOQ:Total Cost*
160 140 120 100

Total Cost
Holding Cost

Cost

80 60 40 20 0 0 500

Order Cost

1000

1500

Order Quantity

EOQ: Optimal Order Quantity*

Optimal Quantity = (2*Demand*Setup Cost)/holding cost So for our problem, the optimal quantity is 316

EOQ: Important Observations*

Tradeoff between set-up costs and holding costs when determining order quantity. In fact, we order so that these costs are equal per unit time Total Cost is not particularly sensitive to the optimal order quantity
Order Quantity 50% Cost Increase 80% 90% 100% 110% 120% 150% 200%

125% 103% 101% 100% 101% 102% 108% 125%

The Effect of Demand Uncertainty

Most companies treat the world as if it were predictable:
Production and inventory planning are based on forecasts of demand made far in advance of the selling season Companies are aware of demand uncertainty when they create a forecast, but they design their planning process as if the forecast truly represents reality

Recent technological advances have increased the level of demand uncertainty:

Short product life cycles Increasing product variety

Demand Forecast
The three principles of all forecasting techniques:
Forecasting is always wrong The longer the forecast horizon the worst is the forecast Aggregate forecasts are more accurate

SnowTime Sporting Goods

Fashion items have short life cycles, high variety of competitors SnowTime Sporting Goods
New designs are completed One production opportunity Based on past sales, knowledge of the industry, and economic conditions, the marketing department has a probabilistic forecast The forecast averages about 13,000, but there is a chance that demand will be greater or less than this.

Supply Chain Time Lines

Jan 00 Design Feb 00 Sep 00

Jan 01 Production Feb 01 Sep 01

Jan 02 Retailing

Production

SnowTime Demand Scenarios

Demand Scenarios
30% 25% 20% 15% 10% 5% 0%

Probability

00 0

00 0

00 0

00 0 16

80

10

12

14

Sales

18

00 0

00

SnowTime Costs
Production cost per unit (C): $80 Selling price per unit (S): $125 Salvage value per unit (V): $20 Fixed production cost (F): $100,000 Q is production quantity, D demand

Profit = Revenue - Variable Cost - Fixed Cost + Salvage

SnowTime Scenarios
Scenario One:
Suppose you make 12,000 jackets and demand ends up being 13,000 jackets. Profit = 125(12,000) - 80(12,000) - 100,000 = $440,000

Scenario Two:
Suppose you make 12,000 jackets and demand ends up being 11,000 jackets. Profit = 125(11,000) - 80(12,000) - 100,000 + 20(1000) = $ 335,000

SnowTime Best Solution

Find order quantity that maximizes weighted average profit. Question: Will this quantity be less than, equal to, or greater than average demand?

What to Make?
Question: Will this quantity be less than, equal to, or greater than average demand? Average demand is 13,100 Look at marginal cost Vs. marginal profit
if extra jacket sold, profit is 125-80 = 45 if not sold, cost is 80-20 = 60

So we will make less than average

SnowTime Expected Profit

Expected Profit
$400,000 $300,000
Profit

$200,000 $100,000 $0 8000

12000

16000

20000

Order Quantity

SnowTime Expected Profit

Expected Profit
$400,000 $300,000
Profit

$200,000 $100,000 $0 8000

12000

16000

20000

Order Quantity

SnowTime: Important Observations

Tradeoff between ordering enough to meet demand and ordering too much Several quantities have the same average profit Average profit does not tell the whole story Question: 9000 and 16000 units lead to about the same average profit, so which do we prefer?

SnowTime Expected Profit

Expected Profit
$400,000 $300,000

Profit

$200,000 $100,000 $0 8000

12000

16000

20000

Order Quantity

Probability of Outcomes
100%
Probability

80% 60% 40% 20% 0%

-3 00 00 0 -1 00 00 0 10 00 00 30 00 00 50 00 00

Q=9000 Q=16000

Revenue

Key Insights from this Model

The optimal order quantity is not necessarily equal to average forecast demand The optimal quantity depends on the relationship between marginal profit and marginal cost As order quantity increases, average profit first increases and then decreases As production quantity increases, risk increases. In other words, the probability of large gains and of large losses increases

SnowTime Costs: Initial Inventory

Production cost per unit (C): $80 Selling price per unit (S): $125 Salvage value per unit (V): $20 Fixed production cost (F): $100,000 Q is production quantity, D demand

Profit =
Revenue - Variable Cost - Fixed Cost + Salvage

SnowTime Expected Profit

Expected Profit
$400,000 $300,000

Profit

$200,000 $100,000 $0 8000

12000

16000

20000

Order Quantity

Initial Inventory
Suppose that one of the jacket designs is a model produced last year. Some inventory is left from last year Assume the same demand pattern as before If only old inventory is sold, no setup cost Question: If there are 7000 units remaining, what should SnowTime do? What should they do if there are 10,000 remaining?

Initial Inventory and Profit

500000
Profit

400000 300000 200000 100000 0

00 00 00 00 0 0 0 00 14 15 00 50 50 65 80 95 11 12 50 0

Production Quantity

Initial Inventory and Profit

500000
Profit

400000 300000 200000 100000 0

00 00 00 00 0 0 0 00 14 15 00 50 50 65 80 95 11 12 50 0

Production Quantity

Initial Inventory and Profit

500000
Profit

400000 300000 200000 100000 0

00 00 00 00 0 0 0 00 14 15 00 50 50 65 80 95 11 12 50 0

Production Quantity

Initial Inventory and Profit

500000 400000
Profit

300000 200000 100000 0

5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000 16000

Production Quantity

Supply Contracts
Fixed Production Cost =$100,000 Variable Production Cost=$35

Wholesale Price =$80 Selling Price=$125 Salvage Value=$20

Manufacturer

Manufacturer DC

Retail DC

Stores

Demand Scenarios
Demand Scenarios
30% 25% 20% 15% 10% 5% 0%

Probability

80 00 10 00 0 12 00 0 14 00 0 16 00 0 18 00 0
Sales

Distributor Expected Profit

Expected Profit
500000 400000 300000 200000 100000 0 6000

8000

10000

12000

14000

16000

18000

20000

Order Quantity

Distributor Expected Profit

Expected Profit
500000 400000 300000 200000 100000 0 6000

8000

10000

12000

14000

16000

18000

20000

Order Quantity

Supply Contracts (cont.)

Distributor optimal order quantity is 12,000 units Distributor expected profit is $470,000 Manufacturer profit is $440,000 Supply Chain Profit is $910,000
Is there anything that the distributor and manufacturer can do to increase the profit of both?

Supply Contracts
Fixed Production Cost =$100,000 Variable Production Cost=$35

Wholesale Price =$80 Selling Price=$125 Salvage Value=$20

Manufacturer

Manufacturer DC

Retail DC

Stores

Retailer Profit (Buy Back=$55)

600,000

Retailer Profit

500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Order Quantity

Retailer Profit (Buy Back=$55)

600,000

$513,800

Retailer Profit

500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Order Quantity

Manufacturer Profit (Buy Back=$55)

600,000

Manufacturer Profit

500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Manufacturer Profit (Buy Back=$55)

600,000

Manufacturer Profit

500,000 400,000 300,000 200,000 100,000 0

$471,900

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Supply Contracts
Fixed Production Cost =$100,000 Variable Production Cost=$35

Wholesale Price =$?? Selling Price=$125 Salvage Value=$20

Manufacturer

Manufacturer DC

Retail DC

Stores

Retailer Profit (Wholesale Price $70, RS 15%)

600,000

Retailer Profit

500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Order Quantity

Retailer Profit (Wholesale Price $70, RS 15%)

600,000

$504,325

Retailer Profit

500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Order Quantity

Manufacturer Profit (Wholesale Price $70, RS 15%)

700,000

Manufacturer Profit

600,000 500,000 400,000 300,000 200,000 100,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Manufacturer Profit (Wholesale Price $70, RS 15%)

700,000

Manufacturer Profit

600,000 500,000 400,000 300,000 200,000 100,000 0

$481,375

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Supply Contracts

Strategy Sequential Optimization Buyback Revenue Sharing

Retailer Manufacturer 470,700 440,000 513,800 471,900 504,325 481,375

Total 910,700 985,700 985,700

Supply Contracts
Fixed Production Cost =$100,000 Variable Production Cost=$35

Wholesale Price =$80 Selling Price=$125 Salvage Value=$20

Manufacturer

Manufacturer DC

Retail DC

Stores

Supply Chain Profit

1,200,000

Supply Chain Profit

1,000,000 800,000 600,000 400,000 200,000 0

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Supply Chain Profit

1,200,000

Supply Chain Profit

1,000,000 800,000 600,000 400,000 200,000 0

$1,014,500

60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0
Production Quantity

Supply Contracts
Strategy Sequential Optimization Buyback Revenue Sharing Global Optimization Retailer Manufacturer 470,700 440,000 513,800 471,900 504,325 481,375 Total 910,700 985,700 985,700 1,014,500

Supply Contracts: Key Insights

Effective supply contracts allow supply chain partners to replace sequential optimization by global optimization Buy Back and Revenue Sharing contracts achieve this objective through risk sharing

Contracts and Supply Chain Performance

Contracts for Product Availability and Supply Chain Profits
Buyback Contracts Revenue-Sharing Contracts Quantity Flexibility Contracts

Contracts to Coordinate Supply Chain Costs Contracts to Increase Agent Effort Contracts to Induce Performance Improvement

Contracts for Product Availability and Supply Chain Profits

Many shortcomings in supply chain performance occur because the buyer and supplier are separate organizations and each tries to optimize its own profit Total supply chain profits might therefore be lower than if the supply chain coordinated actions to have a common objective of maximizing total supply chain profits Double marginalization results in suboptimal order quantity An approach to dealing with this problem is to design a contract that encourages a buyer to purchase more and increase the level of product availability The supplier must share in some of the buyers demand uncertainty, however

Contracts for Product Availability and Supply Chain Profits: Buyback Contracts
Allows a retailer to return unsold inventory up to a specified amount at an agreed upon price Increases the optimal order quantity for the retailer, resulting in higher product availability and higher profits for both the retailer and the supplier Most effective for products with low variable cost, such as music, software, books, magazines, and newspapers Downside is that buyback contract results in surplus inventory that must be disposed of, which increases supply chain costs Can also increase information distortion through the supply chain because the supply chain reacts to retail orders, not actual customer demand

Contracts for Product Availability and Supply Chain Profits: Revenue Sharing Contracts The buyer pays a minimal amount for each unit purchased from the supplier but shares a fraction of the revenue for each unit sold Decreases the cost per unit charged to the retailer, which effectively decreases the cost of overstocking Can result in supply chain information distortion, however, just as in the case of buyback contracts

Contracts for Product Availability and Supply Chain Profits: Quantity Flexibility Contracts
Allows the buyer to modify the order (within limits) as demand visibility increases closer to the point of sale Better matching of supply and demand Increased overall supply chain profits if the supplier has flexible capacity Lower levels of information distortion than either buyback contracts or revenue sharing contracts

Contracts to Coordinate Supply Chain Costs

Differences in costs at the buyer and supplier can lead to decisions that increase total supply chain costs Example: Replenishment order size placed by the buyer. The buyers EOQ does not take into account the suppliers costs. A quantity discount contract may encourage the buyer to purchase a larger quantity (which would be lower costs for the supplier), which would result in lower total supply chain costs Quantity discounts lead to information distortion because of order batching

Contracts to Increase Agent Effort

There are many instances in a supply chain where an agent acts on the behalf of a principal and the agents actions affect the reward for the principal Example: A car dealer who sells the cars of a manufacturer, as well as those of other manufacturers Examples of contracts to increase agent effort include two-part tariffs and threshold contracts Threshold contracts increase information distortion, however

Contracts to Induce Performance Improvement

A buyer may want performance improvement from a supplier who otherwise would have little incentive to do so A shared savings contract provides the supplier with a fraction of the savings that result from the performance improvement Particularly effective where the benefit from improvement accrues primarily to the buyer, but where the effort for the improvement comes primarily from the supplier

Supply Contracts: Case Study

Example: Demand for a movie newly released video cassette typically starts high and decreases rapidly
Peak demand last about 10 weeks

Blockbuster purchases a copy from a studio for $65 and rent for $3
Hence, retailer must rent the tape at least 22 times before earning profit

Retailers cannot justify purchasing enough to cover the peak demand

In 1998, 20% of surveyed customers reported that they could not rent the movie they wanted

Supply Contracts: Case Study

Starting in 1998 Blockbuster entered a revenue sharing agreement with the major studios
Studio charges $8 per copy Blockbuster pays 30-45% of its rental income

Even if Blockbuster keeps only half of the rental income, the breakeven point is 6 rental per copy The impact of revenue sharing on Blockbuster was dramatic
Rentals increased by 75% in test markets Market share increased from 25% to 31% (The 2nd largest retailer, Hollywood Entertainment Corp has 5% market share)

(s, S) Policies
For some starting inventory levels, it is better to not start production If we start, we always produce to the same level Thus, we use an (s,S) policy. If the inventory level is below s, we produce up to S. s is the reorder point, and S is the order-up-to level The difference between the two levels is driven by the fixed costs associated with ordering, transportation, or manufacturing

A Multi-Period Inventory Model

Often, there are multiple reorder opportunities
Consider a central distribution facility which orders from a manufacturer and delivers to retailers. The distributor periodically places orders to replenish its inventory

Reminder:

The Normal Distribution

Standard Deviation = 5

Standard Deviation = 10

Average = 30
0 10 20 30 40 50 60

The DC holds inventory to:

Satisfy demand during lead time Protect against demand uncertainty Balance fixed costs and holding costs

 Normally distributed random demand Fixed order cost plus a cost proportional to amount ordered. Inventory cost is charged per item per unit time If an order arrives and there is no inventory, the order is lost The distributor has a required service level. This is expressed as the the likelihood that the distributor will not stock out during lead time. Intuitively, how will this effect our policy?

The Multi-Period Continuous Review Inventory Model

A View of (s, S) Policy

S
Inventory Position

Inventory Level

Lead Time

s 0 Time

The (s,S) Policy

(s, S) Policy: Whenever the inventory position drops below a certain level, s, we order to raise the inventory position to level S. The reorder point is a function of:
The Lead Time Average demand Demand variability Service level

Notation
AVG = average daily demand STD = standard deviation of daily demand LT = replenishment lead time in days h = holding cost of one unit for one day K = fixed cost SL = service level (for example, 95%). This implies that the probability of stocking out is 100%-SL (for example, 5%) Also, the Inventory Position at any time is the actual inventory plus items already ordered, but not yet delivered.

Analysis
The reorder point (s) has two components:
To account for average demand during lead time: LTAVG To account for deviations from average (we call this safety stock) z STD LT where z is chosen from statistical tables to ensure that the probability of stockouts during leadtime is 100%-SL.

Since there is a fixed cost, we order more than up to the reorder point: Q=(2 K AVG)/h The total order-up-to level is: S=Q+s

Example
The distributor has historically observed weekly demand of: AVG = 44.6 STD = 32.1 Replenishment lead time is 2 weeks, and desired service level SL = 97% Average demand during lead time is: 44.6 2 = 89.2 Safety Stock is: 1.88 32.1 2 = 85.3 Reorder point is thus 175, or about 3.9 = (175/44.6) weeks of supply at warehouse and in the pipeline

Example, Cont.
Weekly inventory holding cost: 0.87= (0.18x250/52)
Therefore, Q=679

Order-up-to level thus equals:

Reorder Point + Q = 176+679 = 855

Periodic Review
Suppose the distributor places orders every month What policy should the distributor use? What about the fixed cost?

Base-Stock Policy
r r

L
Inventory Position

Inventory Level

Base-stock Level

0 Time

Periodic Review Policy

Each review echelon, inventory position is raised to the base-stock level. The base-stock level includes two components:
Average demand during r+L days (the time until the next order arrives): (r+L)*AVG Safety stock during that time: z*STD* r+L

Risk Pooling
Consider these two systems:
Warehouse One
Supplier Warehouse Two Market Two Market One

Market One Supplier Warehouse Market Two

Risk Pooling
For the same service level, which system will require more inventory? Why? For the same total inventory level, which system will have better service? Why? What are the factors that affect these answers?

Risk Pooling Example

Compare the two systems:
two products maintain 97% service level $60 order cost $.27 weekly holding cost $1.05 transportation cost per unit in decentralized system, $1.10 in centralized system 1 week lead time

Risk Pooling Example

Week Prod A, Market 1 Prod A, Market 2 Prod B, Market 1 Product B, Market 2

1 33 46 0 2

2 45 35 2 4

3 37 41 3 0

4 38 40 0 0

5 55 26 0 3

6 30 48 1 1

7 18 18 3 0

8 58 55 0 0

Risk Pooling Example

Warehouse Product AVG Market 1 Market 2 A A 39.3 38.6 STD 13.2 12.0 CV .34 .31

Market 1 Market 2

B B

1.125 1.25

1.36 1.58

1.21 1.26

Risk Pooling Example

Warehouse Product AVG Market 1 Market 2 Market 1 Market 2 Cent. Cent A A B B A B 39.3 38.6 1.25 STD CV 13.2 .34 12.0 .31 s 65 62 S 197 193 29 29 Avg. % Inven. Dec. 91 88 14 15 132 20 36% 43%

1.125 1.36 1.21 4 1.58 1.26 5 77.9 20.7 .27 2.375 1.9 .81

118 304 6 39

Risk Pooling: Important Observations

Centralizing inventory control reduces both safety stock and average inventory level for the same service level. This works best for
High coefficient of variation, which increases required safety stock. Negatively correlated demand. Why?

What other kinds of risk pooling will we see?

To Centralize or not to Centralize

What is the effect on:
Safety stock? Service level? Overhead? Lead time? Transportation Costs?

Centralized Systems*
Supplier

Warehouse

Retailers

Centralized Decision

Centralized Distribution Systems*

Question: How much inventory should management keep at each location? A good strategy: The retailer raises inventory to level Sr each period The supplier raises the sum of inventory in the retailer and supplier warehouses and in transit to Ss If there is not enough inventory in the warehouse to meet all demands from retailers, it is allocated so that the service level at each of the retailers will be equal.

Inventory Management: Best Practice

Periodic inventory reviews Tight management of usage rates, lead times and safety stock ABC approach Reduced safety stock levels Shift more inventory, or inventory ownership, to suppliers Quantitative approaches

Changes In Inventory Turnover

Inventory turnover ratio = annual sales/avg. inventory level Inventory turns increased by 30% from 1995 to 1998 Inventory turns increased by 27% from 1998 to 2000 Overall the increase is from 8.0 turns per year to over 13 per year over a five year period ending in year 2000.

Inventory Turnover Ratio

Industry
Dairy Products Electronic Component Electronic Computers Books: publishing Household audio & video equipment Household electrical appliances Industrial chemical

Upper Quartile 34.4 9.8 9.4 9.8 6.2 8.0 10.3

Median 19.3 5.7 5.3 2.4 3.4 5.0 6.6

Lower Quartile 9.2 3.7 3.5 1.3 2.3 3.8 4.4

Factors that Drive Reduction in Inventory

Top management emphasis on inventory reduction (19%) Reduce the Number of SKUs in the warehouse (10%) Improved forecasting (7%) Use of sophisticated inventory management software (6%) Coordination among supply chain members (6%) Others

Factors that Drive Inventory Turns Increase

Better software for inventory management (16.2%) Reduced lead time (15%) Improved forecasting (10.7%) Application of SCM principals (9.6%) More attention to inventory management (6.6%) Reduction in SKU (5.1%) Others

Forecasting
Recall the three rules Nevertheless, forecast is critical General Overview:
Judgment methods Market research methods Time Series methods Causal methods

Judgment Methods
Assemble the opinion of experts Sales-force composite combines salespeoples estimates Panels of experts internal, external, both Delphi method
Each member surveyed Opinions are compiled Each member is given the opportunity to change his opinion

Market Research Methods

Particularly valuable for developing forecasts of newly introduced products Market testing
Focus groups assembled. Responses tested. Extrapolations to rest of market made.

Market surveys
Data gathered from potential customers Interviews, phone-surveys, written surveys, etc.

Time Series Methods

Past data is used to estimate future data Examples include
Moving averages average of some previous demand points. Exponential Smoothing more recent points receive more weight Methods for data with trends:
Regression analysis fits line to data Holts method combines exponential smoothing concepts with the ability to follow a trend

Methods for data with seasonality

Seasonal decomposition methods (seasonal patterns removed) Winters method: advanced approach based on exponential smoothing

Complex methods (not clear that these work better)

Causal Methods
Forecasts are generated based on data other than the data being predicted Examples include:
Inflation rates GNP Unemployment rates Weather Sales of other products

Selecting the Appropriate Approach:

What is the purpose of the forecast?
Gross or detailed estimates?

What are the dynamics of the system being forecast?

Is it sensitive to economic data? Is it seasonal? Trending?

How important is the past in estimating the future? Different approaches may be appropriate for different stages of the product lifecycle:
Testing and intro: market research methods, judgment methods Rapid growth: time series methods Mature: time series, causal methods (particularly for long-range planning)

It is typically effective to combine approaches.