Many organizations have a hard time determining whether their projects are successful or not. One of the best ways to determine project success is to define a project scorecard before the project starts. The project scorecard identifies a set of metrics and targets that will help to quantifiably state whether the project was a success.
Once this idea is explained, many organizations run out and start to collect measures in many different areas of project performance. However, they soon find out that this is just part of the equation. The collection of metrics information by itself provides only limited value. Most of the value comes when you can compare your actual results against some type of standard or target.
For many metrics, there is an explicit target. For instance, the collection of actual effort, duration, and cost information is used to compare against the estimated effort, duration, and cost that you calculated before the project started. Other examples of these explicit targets include things like costs savings that you are trying to achieve, responding to user requests within two business days, or building your deliverables to a stated level of quality.
For many metrics, however, there is not necessarily an explicit target to attain. For instance, if you track your client's satisfaction with the project team, they may rate your team an average of 3.8 out of a 5.0 scale. However, is that a good number or a bad number? If you're measuring for the first time and you have nothing to compare this with, it's hard to say. Likewise, let's say you count errors that are uncovered in your testing process. If you uncovered 25 errors in your first week, how do you know if that's good or bad?
One way to declare success when you're not sure of the target is to first find a baseline and then look for improvement from that point forward. The baseline is established by taking an initial measurement to see what the current state is. Your target is then based on making some improvement from the baseline.
Let's revisit the client satisfaction survey example mentioned earlier. If you didn't have a valid survey target to shoot for, you can take the initial survey result of 3.8 out of 5.0 as a baseline. Your project scorecard target might then be that you will improve this baseline number by 10% before the project ends. You could then take a similar client satisfaction survey halfway through the project and at the end. If your final survey numbers average 4.18 (3.8 * 1.10), then you can declare success for that measure.
The beauty of this technique is that it works regardless of the initial baseline average. If your first survey was a 3.0 average, then your final survey should be at least a 3.3 (3.0 * 1.10). If your first survey baseline was a 3.3, your final survey numbers should be 3.63 (3.3 * 1.10). The target is a 10% improvement from your first baseline measure, regardless of what that initial average number is.
When you're trying to objectively measure the success of your project, it's imperative that you have good targets in place. Many targets can be stated with some degree of confidence. Other are not known ahead of time. When you don't know what a good target level is, you can usually measure early to gain a baseline, and then focus on some type of improvement by the time the project is over.