There’re so many tools or methodologies to manage project. But they are not sufficient to judge whether the project goes well. Totally integrated data collection, measurement, analysis, and process would make it better. First of all, monitoring the trend of customers and interpreting it are the preceding task. (But never reveal them to the public as your data collection should keep their trust) That analysis can help estimate the project objectively trusted. Second, communication matters on the project while business is separated development or product line. (or vendor to customers) If one of them are bottle neck, then whole project would be delayed. Thus you’d rather share all the things of the project in your company so that project members can catch the narrow position. Lastly, you should avoid the following situations:

• Ambiguous or diffuse project ownership.
• Stakeholder complexity
• Conflicts of interest
• Inconsistencies (hardest to discern in advance)

Project management is not the easy one and the failed project is desperate. The best is not to start the project if it has less possibility or is in a bad condition.

Requirement is a kind of communication. Properly documented requirements help reduce the gap between stockholders (who mostly require). And understanding requirements from the start synchronize development group with business users and their needs. The following is the list of proper solutions.

• Document the requirements clearly.
• Set priority of each requirements.
• Record the trace of the requirements.
• Technical team should communicate to business users.
• If there lies any change, be clear about the implications for scope and cost.

These rules are not the essential conditions to get the project succeed. But they’re some of guidelines which can avoid project getting down into failure.