1 1 Supply Chain Strategies Case: 7-Eleven Japan OPERATNS 476 Distribution and Supply Chain Management Fall Term 1, 2007 Jeannette Song Session 2 2 Plan •Review of last class •Supply chain strategies •What is the right supply chain for your product? •What kind of supply chain partner would you like to choose, as a supplier and as a customer? •Review: Newsvendor model •Case: 7-Eleven Japan
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
1
Supply Chain StrategiesCase: 7-Eleven Japan
OPERATNS 476Distribution and Supply Chain Management
Fall Term 1, 2007
Jeannette Song
Session 2
2
Plan
•Review of last class
•Supply chain strategies
•What is the right supply chain for your product?
•What kind of supply chain partner would you like to choose, as a supplier and as a customer?
•Review: Newsvendor model
•Case: 7-Eleven Japan
2
3
Review of Session 1•What is a supply chain?
•Products, value added activities/ tiers, facilities, players, flows of materials, information and cash
•What is supply chain management?•Purpose & ideal•Elements•Challenges•Some principles and innovations
•Example: Supply chain of a bottle of Listerine
•Case: Barilla•VMI (Vendor Managed Inventory; also known as Continuous Replenishment)
•Why and how?•Collaboration and incentives: external and internal
4
Review of Inventory Concepts•Demand per period: mean λ, standard deviation σ; replenishment leadtime: L•IP = inventory position; IO = inventory on-order•IN = net inventory = on-hand inventory – backorders•IP = IN + IO (definition)•If full backlogging, all demand will be satisfied eventually,
•E[IO] = average pipeline inventory = λ L = mean leadtime demand•This is Little’s Law: WIP = (Processing Time) * (Throughput Time)
•Base-stock policy with base-stock level s•IP = s•s = E[IN] + E[IO] = E[IN] + λL. So E[IN] = s – λL
• Each year, 95% of the products are completely new designs
• Demand forecasts err by as much as 200%
• Short retail season
10
Physically Efficient Supply Chains
• Cost concerns:– Inventory holding, transportation, handling and
manufacturing costs• What do we usually see?
– Large manufacturing batches to achieve scale– Minimum order quantities to reduce handling/order
processing costs– Full truck load shipments
• Focus on traditional cost measures:– Inventory turns, factory and transportation utilization
6
11
Market Responsive Supply Chains• Cost concerns:
– Inventory obsolescence/markdown, lost sales/poor service
• What to do:– Accept and reduce demand uncertainty:
• Finding sources of new data to learn more about demand• Postpone variety as much as possible, component commonality
– Avoid uncertainty: Pay a premium for flexibility: faster transportation, lower transportation and factory utilization, smaller batches
– Hedge against uncertainty: Buffer inventory and excess capacity.
• Focus on flexibility:– Avoid emphasis on traditional explicit costs (inventory holding,
transportation efficiency, manufacturing utilization)– Recognize that opportunity cost of lost sales and poor service are
high
12
Matching Supply Chains With Product Demand Characteristics
Supply Chain Types
Efficient
Responsive
Functional Innovative
Product Demand
Mismatch
Mismatch
7
13
Paul Pressler's Fall From The Gap…. Despite Pressler's expertise and comfort with the operational
side of the company, some of his initiatives in that area also backfired. In 2003, he kicked off a plan to remake the retailer's supply chain. One goal was to combine fabric purchases across all three of the company's brands, which had previously used separate suppliers. Pressler wanted to cut costs and speed the flow of merchandise from drawing board to store shelf.
But the effort stumbled. Since all three brands sell jeans made of denim, in late 2003 they began purchasing it together. But the big buying deals yielded denim of a single quality, weave, and weight. That meant all three brands were limited to one kind of fabric even though, with their different audiences, they needed to be able to offer an array of jeans. And with all that denim ordered in advance, there was less incentive for designers to keep their eyes open so they could jump on the next great style.Jeans made with material from the pooled purchases ended up largely falling flat, and the joint purchasing effort was abandoned …..
FEBRUARY 26, 2007
14
Samsung’s European Supply Chain
Share the same logistics systems?Same shipping schedule?
8
15
So Many Choices!
Does the world need 28 kinds of toothpastefrom each manufacturer?
16
Supply Chain Strategy
What is the Right Supply Chain for My
Company?
Products
9
17
How Do You Choose Supply Chain Partners?
• Suppliers
• Customers
18
ReviewThe Newsvendor Model
One-time decision about how much (Q) to order to cover demand in a single period
Demand (D) is uncertain, has a distribution F with mean E[D]= λ and standard deviation σ
Inventory remaining at end of period perishes (or is salvaged)
D = demand for newspapers --- distribution FQ = # of newspapers to purchase from distributorp = price per newspaperw = cost per newspaperv = salvage value
Profits are …
( )},min{ },min{),( DQQvQwDQpDQ −+−=π
quantity sold =minimum between demandand quantity available
costs salvage forunsold units
20
Marginal Analysis
Some notation…
G = marginal revenue for buying and selling one unit ( = p – w )
L = marginal loss for buying and not selling one unit ( = w - v )
• Expected marginal benefit of stocking one more unit:
• Expected marginal cost of having one more unit:
( ) )( )(1 QDPGQFG ≥=−
)()( QDPLQFL ≤=
11
21
Optimal Solution
( ) )()(1 ** QLFQFG =−
marginal benefit = marginal cost
Optimal solution:
vpwp
LGGQF
−−
=+
=)( *
22
Normal Demand
Demand (D) is has a normal distribution F with mean E[D]= λ and standard deviation σ
Compute ratio = G/(G+L)
Set z* = NORMSINV(ratio)
Then, the optimal order quantity is
Q* = λ + z* σ= mean demand + safety stock
12
23
Example: Greeting Cards
A greeting card retailer is deciding what quantity of Mother’s Day cards to purchase.
– Purchase price (from distributor) = $1 per card– Retail price = $2.50 per card– Unsold cards are sold to a liquidator for $0.25 per
card– Any excess demand is lost– Based on historical data, the retailer estimates that
demand for the cards will follow a Normal distribution with a mean λ = 200 and standard deviation σ = 50
24
Solution
G =$2.50 – $1 = $1.50L = $1 – $0.25 = $0.75
ratio = 1.5/(1.5+0.75) = 0.6667
safety factor z* = normsinv(0.6667) = 0.45
order quantityQ* = 200 + (0.45)(50) = 221.54
Safety stock = Q*- λ =221.54-200 = 21.54(extra units ordered above the mean demand)
13
25
Case: 7-Eleven JapanLearning Objectives
• Distinguish – functional and innovative products– efficient vs. responsive chains
• How to build a responsive supply chain– Speed – Flexibility– Efficiency?
• Role of information technology• Role of partnership and incentives• Culture; continuous improvement; innovation