Should SPLE be started at all?
- Big upfront investment
- Reorganization, resistance, big change for a company
How will product line engineering impact our business performance?
- What we will gain or lose
What products part of product line?
What shall be the characteristics or features of these products?
- Core asset of the domain
- Specific for a single product
Somebody: Isn't SPLE about reusing?
Reference architecture; then be able to select proper themes. Might have different SPL for different lines of product inside a company while sharing some features between the product lines.
What timing shall be chosen for the market entry?
- You sell products, not product line...
How shall products be produced?
- Which functionality is commonality?
- Which functionality is project specific?
How shall we introduce PLE in our organization?
- Maybe one department makes reqs, another testing, anothes develops
- Good to restructure
- One team on the domain
- Other teams on the products
- Depends...
How shall we evolve PL over time?
- Evolve the platform and the single products
The answer is related to the economics of the market...
One fundamental part is the business aspect. What you gain from going SPL.
Strategy
-
How new products are defined
-
Customer-driven
-
Mass customization. Lots of variants => additional cost. Companies don't know what the customers want, Huawei has this problem. Don't know if the customers want or use the apps. We only use a subset of the functionality, might not need all features.
Imagine the software on a Volvo. Might be features that we want and some we don't.
-
-
Product-driven
-
Market-oriented. Commonly: One leader company and many followers. Risky to be the leader of a innovative product => might fail => 2nd or 3rd might more easily succeed.
-
Technology-oriented.
-
-
-
Market strategies, two different aspects
-
Porter categorization
- Aims at providing the lowest possible costs
- Differentiation
- Focus on a special niche of people. Small subset of the markets. If good SPL then can target many small subsets. Target huge number of niches instead of one product for a large market. Long tail.
-
Kotler categorization
- Improve quality compared to competitors; costs;
- Newer. Company strong on innovation
- Faster. New possibilities arise quickly
-
-
Product line life-cycle
- Introduction
- Growth
- Maturity
- Saturation
- Degeneration. Start to think about changing products or a new products that can substitute the current one
-
Strategy
- Reduce the amount of effort (and then cost) and time to market. To substitute a product we need to understand the market and know when saturation has been met
- Quality benefits of PLE usually plays a secondary role.
-
Product line economics
- Product qualities
- Process qualities
PLE reduces development risks. Knows that the core assets already works.
- SIMPLE Model of Product Line Economics. http://simple.sei.cmu.edu
- Reorganization
- Process improvement
- Training
These's always variability.
External variability visible to the customer.
Internal variability is domain artefacts that's hidden to the customer.
Part of the core assets is commonality. E-shop example. Depencency notaition.
http://splot-research.org Create a Feature model for the software of an e-commerce webside. Focus on external variability. Prepare a presentation for Monday 15:15.