When the software project starts one of the critical things is project deadline. But in most cases deadlines come and go, come and go. And as the deadline is missed, it is moved a week or two ahead, but that actually doesn't change anything. That is why we all need realistic schedules and stuff from the management like "CEO said we must build the tool until February 12th" or "our competitor releases in a week, we MUST finish in 5 days" simply doesn't work. You can tighten schedules as much as you want, it won't help. You need a real project schedule (sometime inside the project teams there actually exists two types of schedules: one for the upper management, to make everything look nice, and the other for team members).

It is very hard to define the exact schedule for a software development product, there too many risks to be taken into account, so what can we do to get the best results?

• Set the deadlines and milestones as the project moves ahead. Yes I know the classic software development approach says that all the planning and estimating has to be done before the programming (implementation) starts. I personally think that agile software development approach is the most realistic. Project is conducted in small iterations (1-4 weeks) , where each iteration is an entire software project: including planning, requirement analysis, design, programming, testing, documentation. It is easy and effective to manage your iterations with Comindwork, you have everything you need: wiki, task management, milestones, and all other features available online 24/7.
• Trusted and reliable project team. First of all - all team members should be allowed to estimate the tasks, related directly to them. Secondly - from marketing, to documentation team, an active participation in project processed is a necessity if we want the most realistic schedules. This brings more confident and responsibility to the team.
• There are no identical projects. Every project by itself is unique, and it's a common problem when the managers and project members estimate the new project as it is the "same" as some other project. E.g. if you have create a website using .Net it doesn't mean that you can easily estimate any site that needs to be developed under .Net. Experience in the similar projects is a great of a value for the project team as long as it is not used for 100% comparisons and estimations. Comindworks' colloborative environment will help you achieving  this.
• Work closely with the customer. Effective communication with customers is one of the project success factors. It is ideal when customer participates in product testing, provided notes are very valuable, because they can actually point out bugs that are critical. Usually developers and customers evaluate bugs differently. Close collaboration with customers of the project will help you to avoid developing features that will never be used. Comindwork makes a perfect assistant here, especially if your team is distributed and needs to collaborate real time online.

Unreal schedules forces project teams to waste time, valuable time that could be spent finishing the software. Only when project members, prior project start, will recognize that the projects are hard and that it is hard to create accurate schedules, only then the project can be managed and developed in a realistic way. F.Brooks addressed this issue year ago in his paper No Silver Bullet, and it is still actual today. As well you can read more about realistic scheduling in a software development project here (article from IBM). What techniques do you use? Share your tips with us :)