The site specification is the planning team's concise statement of core goals, values, and intent, to provide the ultimate policy direction for everything that comes next. Designing a substantial Web site is a costly and time-consuming process. When you're up to your neck in the daily challenges of building the site, it can be surprisingly easy to forget why you are doing what you are, to lose sight of your original priorities, and to not know on any given day whether the detailed decisions you are making actually support those overall goals and objectives. A well-written site specification is a powerful daily tool for judging the effectiveness of a development effort. It provides the team with a compass to keep the development process focused on the ultimate purposes of the site. As such, it quickly becomes a daily reference point to settle disputes, to judge the potential utility of new ideas as they arise, to measure progress, and to keep the development team focused on the ultimate goals.
At minimum, a good site specification should define the content scope, budget, schedule, and technical aspects of the Web site. The best site specifications are very short and to the point, and are often just outlines or bullet lists of the major design or technical features planned. The finished site specification should contain the goals statement from the planning phase, as well as the structural details of the site.
These are big questions, and the broad conceptual issues are too often dismissed as committees push toward starting the "real work" of designing and building a Web site. However, if you cannot confidently answer all of these questions, then no amount of design or production effort can guarantee a useful result.
The site specification defines the scope of your project that is, what and how much you need to do, the budget, and the development schedule. "Scope creep" is the most prevalent cause of Web project failures. In badly planned projects, scope creep is the gradual but inexorable process by which previously unplanned "features" are added, content and features are padded to mollify each stakeholder group, major changes in content or site structure during site construction are made, and more content or interactive functionality than you originally agreed to create is stuffed in. No single overcommitment is fatal, but the slow, steady accumulation of additions and changes is often enough to blow budgets, ruin schedules, and bury what might have been an elegant original plan under megabytes of muddle and confusion.
Don't leap into building a Web site before you understand what you want to accomplish and before you have developed a solid and realistic site specification for creating your Web site. The more carefully you plan, the better off you will be when you begin to build your site.
One excellent way to keep a tight rein on the overall scope of the site content is to specify a maximum page count in the site specification. Although a page count is hardly infallible as a guide (after all, Web pages can be arbitrarily long), it serves as a constant reminder to everyone involved of the project's intended scope. If the page count goes up, make it a rule to revisit the budget implications automatically the cold realities of budgets and schedules will often cool the enthusiasm to stuff in "just one more page." A good way to keep a lid on scope creep is to treat the page count as a "zero sum game." If someone wants to add pages, it's up to them to nominate other pages to remove or to obtain a corresponding increase in the budget and schedule to account for the increased work involved.
Changes and refinements can be a good thing, as long as everyone is realistic about the impact of potential changes on the budget and schedule of a project. Any substantial change to the planned content, design, or technical aspects of a site must be tightly coupled with a revision of the budget and schedule of the project. People are often reluctant to discuss budgets or deadlines frankly and will often agree to substantial changes or additions to a development plan rather than face an awkward conversation with a client or fellow team member. But this acquiescence merely postpones the inevitable damage of not dealing with scope changes rationally.
The firm integration of schedule, budget, and scope is the only way to keep a Web project from becoming unhinged from the real constraints of time, money, and the ultimate quality of the result. A little bravery and honesty up front can save you much grief later. Make the plan carefully, and then stick to it.