|A big and visible project can be daunting indeed. Where do you begin? I recommend starting with questions, and the importance of thinking through what you are trying to answer before you talk about requirements and potential solutions. It was Rilke who taught me the value of learning to love the questions.
I remember a humanitarian organization project where the CEO had seen an Emergency Operations Center (EOC) in another organization. He wanted one for ours. And he wanted it fast. So we formed a project team of the best people from the business unit (Disaster Preparedness), the IT department and a consultant partner. We kicked the project off by meeting with the CEO, then the SVP of the Division and then the Head of the department. These were the business owners. Rather than the traditional needs analysis approach, we asked each of them "what questions can't you answer today or are difficult to get an answer that this system should be able to answer?"
The CEO and SVP had four questions:
1. What are the details for this EVENT? (Are we matching size of response to size of disaster?)
2. What are the ASSETS and resources we can bring to bear?
3. What PROGRESS and decisions are being made?
4. How are we INFORMING others?
The Head of Department had 14, more detailed questions, which I won't list here. What was apparent was the closer to the work, the greater the demands. No surprise. The project team posted these questions on the EOC wall and referred to them often, from the beginning.
Framing the questions is a good way to start thinking about technology projects. As with strategy, challenge your business project sponsors to think about the questions that they cannot answer today, or find difficult to answer, that the application will make easier to answer. Coming up with the 10-20 questions to answer will help focus the project (and structure the demo for the CEO). It will also provide a more interesting and relevant acceptance test. If the new system can answer the questions, you’ve arrived. 
This is probably the most important exercise of IT project management.