出自Stéphane DEPRES – 定义架构路线图，定义项目的业务价值，使用看板拉式系统方法管理项目组合，建立足够的指示器。。。12个管理你的项目组合的最佳实践。
Everybody should agree that the definition of an IT strategic plan is fundamental in an IT organization because it aligns the IT with the business strategy.
In most organization, this IT strategic plan includes a project roadmap trying to address business strategic needs. Of course this project roadmap cannot be achieved without defining an IT architecture roadmap: software architects should be involved to analyze project interdependencies related to application decommissioning, interfaces changes, and frameworks or common components development.
Because some projects have real business deadline constraints, because there are project interdependencies, many organization schedule projects several years before starting. A vision at this point is great; a commitment is much too early. Firstly effort estimation before the project has even started should be considered as wrong. Secondly the available resources and budget will perhaps not be the one forecasted.
After the project start, even if there is a committed delivery date, the real life is sometime different. And a project that is late or need more budget than expected will impact other projects…
It means that project portfolio should be really managed! It should be used to make decisions, tradeoff and assignments taking into account the whole system complexity including project interdependencies and constraints.
There are several guidelines that you should follow when managing your project portfolio in order to achieve your strategy and minimize business and project teams' frustration:
Align the project portfolio with your real budget / organization capacity. Be ambitious but realistic.
You define business value for the user stories of your Agile projects. Do the same for your projects!
What is this Project Return Of Investment? What are the risks? Should we do this project at all? It means that you should:
- define the business value of your projects,
- prioritize your project based on their business value or based on their business value / cost ratio (but of course you should also take into account the project interdependencies),
- do not plan a project with a too low business value or a two low ratio.
Define architecture roadmap and analyze project interdependencies in involving architects.
Manage project portfolio using a Kanban pull system approach:
- do not start too many projects at the same time,
- wait current projects are finished before starting new ones,
- allocates enough resources to the running projects.
Little law demonstrates that projects duration will decrease. Project interdependencies problem should also decrease.
Use Agile methodology on your projects! Agile has many advantages:
- it allows delivering products that best suit the business needs,
- it provides value to the business at the earliest,
- it allows more flexibility relative to requirement changes,
- it allows delivering on time even if scope should be adjusted (as user stories with high business value are implemented first, only cosmetic user stories shall remain at the project end. If the project is late, last user stories can then be removed from the scope without major business impact).
But Agile do not mean no visibility! Require your projects to provide a multi-releases burndown chart to provide visibility on the forecasted releases delivery dates.
Before a project start ensure that all stakeholders (business and IT) share the same project vision (outline of the envisioned key needs and features, business processes that should be changed, change management required, key stakeholders). Pay attention to the project start process and communicate appropriately.
It should also be clear for all stakeholders that this vision should be refined during the first project phase (scoping phase, iteration 0…). In particular the project scope, the budget, the project duration and the resources shall be aligned at the end of this phase.
Follow-up the project portfolio regularly and take only decision when there are decision to take!
If there is really the need to take a decision:
- take it clearly in analyzing all the impacts,
- communicate with the decision rational.
It is also important that the project portfolio address project risks and issues and associated mitigation and corrective actions. Some decisions and actions are outside of the project responsibilities and should be addressed by the appropriate committee.
Be very cautious when you define indicators to manage. Every indicator should have a clear objective and should not be ambiguous. Document its specification, how it will be analyzed, how data should be collected and communicate it.
You should also know that each indicator you define will have a positive and a negative impact in your organization. For instance if you decide that each project should manage an "on budget" indicator:
- positive impact: project managers will follow the budget and will take corrective actions if needed
- negative impact: some project manager may overestimate the project cost
Be very cautious too when you analyze indicators: there can be several factors leading to a given status. So it's important to analyze all indicators before taking a decision. And of course it is necessary to heard project managers explanation!
Create a positive, win-win culture in your organization. Do not blame project team for being late. Try to find solutions!
Defining an IT strategic plan, managing a project portfolio is not easy! But now you have some keys….