Product Family and Product Portfolio design Product Portfolio

Product Family and Product Portfolio design Product Portfolio A set of different products offered by a company at a given point of time. Product architecture A hierarchical, structural description of each product, Often based on functional decomposition Product Portfolio Architecture Description of components/modules/systems of a set of products (past, present, future) offered by the company

Product Family and Product Portfolio design Different products do not share any components or modules [High sales volume]

Product Family and Product Portfolio design Products do not share components, Have adjustable input parameters

Product Family and Product Portfolio design series of products that share some modules (software, hardware) Images (c) Krups

Product Family and Product Portfolio design Each subsequent model enhances only some subset of the modules of the previous generation (a) Canon® IXUS 700, Feb 05 (b) Canon® IXUS 750, Aug 05 Same CCD, Lens; Different LCD, controls

Product Family and Product Portfolio design Architecture differentiates modules based on “module is consumable” or not Examples: Ink-cartridges for ink-jet printers film for film cameras, …

Product Family and Product Portfolio design ATI Radeon X 300 (LE) Nvidia Ge. Force PCX 5750 Interface (mech, elec, software) follows a published standard Canon EF-lens mount system Bus-architecture in PC’s

Product Family and Product Portfolio design MICRON 1 GB DDR 2 533 MHZ DIMM MODULE MICRON 256 MB DDR 2 533 MHZ DIMM MODULE modules may be parametrically modified and fit into the product at the time they are ordered Examples: PC modules

Product Family and Product Portfolio design Product designed and made Based on customer-provided Engg spec Example: Tsing Ma Bridge

Product Family and Product Portfolio design Customer configures product by selecting a series of different modules that make up the product Dell PCs, Automobiles, Swatch, My. Barbie, NIKEi. D, …

Designing Portfolios with multiple products 1. Statistical method: the customer-market model Interview customers identify reqts SURVEYS: Importance of each reqt Identify target values (m, s) for each reqt Identify time-variations of reqts Segregate customer-base into segments (factor analysis) Analyze specs for each segment for each requirement m 1 m 2 … mk mpop s 1 s 2 … sk spop

Designing Portfolios with multiple products 1. Statistical methods: demand assessments via customer surveys Possible decision tree N m constant over time? modular generations Y spop is large ? N fixed unshared portfolio Y segment s’s ≈ population s ? N modular families Y parametric/adjustable portfolio (e. g. car seats)

How to design Portfolios with multiple products 2. Functional Architecture: Break the product into functional modules power saw screwdriver scumbuster

How to design Portfolios with multiple products 2. Functional Architecture: Break the product into functional modules

How to design Portfolios with multiple products 3. Systematic Design Approach for size-ranging (Pahl and Beitz) Advantages of scaling: (a) The products are geometrically similar (b) We may use the same design drawings/models (change scale) (c) Manufacturing benefits: e. g. pantographs, tooling design Pahl and Beitz: Optimal schemes use geometric scaling

Systematic Design Approach for size-ranging. . Scaling and Non-dimensional techniques Length of a feature in the base design = L 0, Length of the same feature in a different size of model = L 1 Two products are geometrically similar if, for each feature, the ratio f. L = L 0 / L 1 f. L is non-dimensional

Systematic Design Approach for size-ranging. . Scaling and Non-dimensional techniques Step 1. Find the functional characteristic Step 2. Find a non-dimensional constant related to the characteristic Step 3. Define size steps for functional characteristic in geometric series Step 4. Scale all geometric features to maintain same non-dimensional constant value (functionality). heat dissipation = K ( surface area) scale factor f. L area increases by f. L 2 increase in heat generation = K(f. L ) A simple heat sink => heat sink may be scaled up by f. L 0. 5

How to design Portfolios with multiple products 4. Taguchi’s Loss Function Men’s dress shirts: 10 sizes Each size: neck (collar length) and sleeve (arm length). neck sizes increments: 1 cm sleeve lengths increments: 2 cm. Person size ≠ nearest shirt size “Loss of Quality”

4. Taguchi’s Loss Function Men’s dress shirts Neck size = y Target = m If y ≠ m, there is a loss Loss = L(y) = L( m + (y-m)) Using Taylor-series expansion: L(m) + (y-m) L’(m)/ 1! + (y – m)2 L”(m)/ 2! + …

4. Taguchi’s method: Loss function. . Loss = L(y) = L( m + (y-m)) = L(m) + (y-m) L’(m)/ 1! + (y – m)2 L”(m)/ 2! + … Ideally: (a) L(m) = 0 [if actual size = target size, Loss = 0], and (b) When y = m, the loss is at its minimum, therefore L'(m) = 0 Taguchi’s Approximation: L(y) ≈ k( y – m)2

4. Taguchi’s Method: Loss Function… Assume tolerance limits : D 1 and D 2 neck size in range: [ m – D 1, m + D 2]. out-of-tolerance losses: D 1 and D 2 Assume over-size/under-size loss = $ 400 L(y) = k( y – m)2 y < m, k 1 = 400/ D 12 y > m, k 2 = 400/ D 22. [out-of-limit]

4. Taguchi’s Method: Loss Function Example Let D 1 = 0. 5 cm, and D 2 = 1 cm, k 1 = 400/ 0. 52 =1600, and k 2 = 400/ 1. 02 = 400, loss function : If a person has a neck size = 40. 2, Or (i) Buy a size 40, => Loss of value = 1600 ( 40. 2 – 40)2 = $64, (ii) Buy a size 41, => Loss of value = 400 ( 41 – 40. 2)2 = $256.

4. Taguchi’s method: determination of size ranges Break even point for customers: 1600( y – 40)2 = 400 ( 41 – y)2 y* = 40. 33 Net loss to community: If manufacturer offers only even sizes (40, 42, 44, …): Break even point: 1600( y – 40)2 = 400 ( 42 – y)2, or y* = 40. 67 Net loss to community:

4. Taguchi’s method: determination of size ranges Break-even model: Compare total consumer loss with manufacturer benefit/order Estimate of manufacturer benefit: (retail price – production cost) Example: (retail price – production cost) = $180 Consumer loss on size ranges (40, 41, …) = $59. 3 < $180 Consumer loss on size ranges (40, 42, …) = $237 > $180 Break even point: steps of 1. 8 cm

How to design Portfolios with multiple products 5. Systematic, Intuitive Methods 5. 1. SCAMPER 5. 2. Orthographic analysis Step 1. List independent attributes Step 2. Coordinate system, one axis per attribute Step 3. List values along each axis Step 4. Combine, Permute, Interpolate, Extrapolate

Orthographic analysis: Food Manufacturer Example current products current clients frozen vegetables frozen mashed potatoes department stores restaurants fried potato chips department stores Fry

Orthographic analysis: Food Manufacturer Example Fry Combine (processes): “fry” + “mash” hash-browns Combine (materials): potato + vegetable vegetarian patty Permute: fry vegetables vegetable chips Extrapolate: add fruits banana chips Interpolate: half cooked potato chips refry-to-eat