PERFORMANCE ANALYSIS

Many of the measures discussed so far could be classed as general performance measures. To enable you to get the best possible information from these measures you need to examine how you might analyse the results of the measurement. There are several techniques for looking at the various areas:

• comparison;

• ratio analysis;

• benchmarking analysis;

• historical analysis.

Comparison

Comparison is perhaps the simplest method that can be applied. It simply means comparing the actual result achieved against the baseline plan. This simple technique enables the project manager quickly to identify where problem areas are. To be able to use comparison techniques on an advanced project effectively nearly always requires some level of automation. Automation is required simply to overcome the large volume of tasks that exist.

Perhaps the simplest method of automating comparison is simply to subtract the actual result from the baseline plan. This can be simply achieved by setting up a table in a spreadsheet program; this is illustrated in Table 6.2.

Table 6.2: Subtraction table

Change request title	Actual	Baseline	Difference
Task 1	5	10	5
Task 2	7	3	4
Task n	x	y	x-y

This table is simply created. After it is calculated it is an easy process to sort the results in ascending order. This then gives a list of the biggest variances from the baseline. Subtraction in this manner can be applied wherever there are two data sets. The key is to identify the baseline set and monitor against it. All the areas of the project can be analysed in this manner.

The comparison method can be particularly useful when trying to understand what is going wrong. Analysing data in a localized manner is very often revealing . Analysis like this is especially effective when carried out in an automated manner. Ideally a project manager should narrow down a problem area and then put in place localized monitoring.

Ratio analysis

Comparison of a data set has its limitations. In particular it doesn't allow for the comparison of tasks of dissimilar size .

Table 6.3 shows three tasks of varying durations. Using the comparison technique it would seem that the task to be monitored is task 2. This may be true. It has the largest number of outstanding days. However, this doesn't allow for a reasoned task comparison. To allow a more reasoned approach, ratios are required. There are several possible ratios:

Table 6.3: Subtraction table tasks of different sizes

Change request title	Actual	Baseline	Difference
Task 1	5	20	15
Task 2	10	100	90
Task 3	5	50	45

	Task 1	Task 2	Task 3
Difference:Actual	15:5 = 3	90:10 = 9	45:5 = 9
Difference:Baseline	15:20 = 0.75	90:100 = 0.9	45:50 = 0.9
Actual:Baseline	5:20 = 0.25	10:100 = 0.1	5:50 = 0.1

From the ratio analysis it's clear that task 1 is the most advanced. The analysis also shows that task 2 and task 3 are equally advanced. Tasks 2 and 3 actually are at a similar point, something that comparison alone would have failed to show.

Benchmarking analysis

Benchmarking analysis is simply a term for comparison of an actual result against something similar. As with comparison and ratio analysis, benchmarking analysis can be used in a variety of ways. The most obvious way of using benchmarking in performance analysis is through analogy analysis. This simply means finding the best practice carried out for a similar activity in industry and comparing the project task against it.

Benchmarking sounds a simple enough task but it can prove hard, especially on advanced projects. Advanced projects as their name suggests tend to be new and untried. This in turn means there are few data from which to create a benchmark. To overcome this difficulty the simplest way forward is to set performance guesses. This is done by gathering as much evidence as possible from similar tasks and then guessing what value to use.

Benchmarking however is not only about technical analogy analysis. It is also about the individuals within the project. When examining performance one of the key success factors is staff. Therefore one of the best measures that can be taken is an analogy assessment of staff. This assessment should examine similar activities that were successful and should look at the staff involved. The assessment should cover:

• staff experience;

• staff skill level;

• staff seniority .

Historical analysis

Although benchmarking is a useful technique it doesn't account for the culture of the organization. At the simplest level, the processes used by the organization using the benchmark and the organization creating the benchmark may be very different. This could easily make the benchmark invalid. To overcome this, historical analysis can be used.

Historical analysis examines previous work of a similar nature. It is similar to analogy analysis but normally focuses specifically on data from the sponsoring organization. Focusing on organizational data in this way means that account is taken of the vagaries of the organization. This means the culture, the people and the operating method are all accounted for. Historical data normally have to be tracked down within an organization. A few organizations are however advanced enough to have stored the data ready for future projects.

Historical performance analysis is carried out by using the comparative analysis or the ratio analysis technique. However, unlike in the previous discussions, the only data that would be used are valid organization data. Generally the project manager can have confidence in the results.

There are many ways of examining historical data, too many to examine in this book. However, some basic points apply to them all:

1. Understand what is different about the historical project and the one being undertaken.

2. Understand the changes that have happened in the organization since the historical project, for example whether new processes have been introduced.

3. Ask questions about trends rather than specific details. With historical data this works better since you cannot understand fully the reasons that caused any particular problem.