HOW TO CALCULATE THE PSI

The PSI is calculated by applying a score to each of the Ten Steps as they relate to the particular project. The maximum possible scores are as follows :

STEP	SCORE	TOTALS
1	20
2	20
3	10
4	10
5	10	70
6	10
7	10
8	10
9
10		30
	Grand total	100

Just as an aside, notice how critical the planning stage is. I am firmly convinced that the success or failure of projects is determined in the first 10 percent of their lifetime. OK, back to the PSI. Here's how you determine each of the scores.

1 Goal {20}

Write down:

• A one- sentence description of the project [8]

• Three or four bulleted items on what constitutes project completion [6]

• A 2 “3 page blurb which attempts to answer the questions on the visualization checklist (see p. 11) [6]

Having done this, score the project using the numbers in square brackets.

2 Job list {20}

• Is the job list up to date? [4]

• Is it complete? Does it cover at least all the items in the checklist on p. 19? [4]

• Are all the major milestones indicated and well-defined ? [6]

• Is it detailed to first milestone? [6]

Score using the numbers in square brackets.

3 One leader {10}

Questions like the following should help you to smoke out this one.

• Name the project leader.

• Is there a person who has the "fire in the belly" to get the project done?

• How many other projects does he or she lead?

Score as follows:

¢ 1 leader	10
¢ 2 leaders	4
¢ 0 or more than 2	1

4 Assign people {10}

Use Form 2 from Case Study 5 (Chapter 4) and the job list from 2 above to score all of the people-jobs in the job list. Add up all the scores and divide by the number of job-persons. Then depending on where the resulting number lies score from 10 as follows:

¢ 1.00 “1.25	10
¢ 1.25 “4.49	4
¢ 4.50 “5.00	1

You could do this in two cuts, first at the person level (divide by the number of people), then at the job-person level.

5 Margin for error {10}

The following tasks will help you to assess the margin for error:

• Write down the major risks

• Describe the fallback position

• Explain how, by differing from the final goal, this fallback position creates for you a margin for error

You could do this at a number of levels:

• At the project level

• For the major milestones

• For those items on the critical path

• For each job-person

Lose 15 from your cumulative score if you have no margin for error.

6 Leadership style {10}

Do the analysis required by Form 2 from Case Study 5 (Chapter 4). Compare with what's happening on the ground. Score out of 10.

7 Know what's going on {10}

Analyze the reporting and monitoring mechanisms in use, and score out of 10. Lose points for no monitoring and controlling against the plan.

8 Tell people what's going on {10}

Analyze information dissemination mechanisms, for example, does everyone have an up-to-date copy of the plan; do they get it each time it changes? Lose points for no progress meetings and no progress reports .

9 Repeat 1 through 8 {No score}

10 The prize; The reckoning {No score}

EXAMPLES

Let's look at some examples. Two, which we know well by now, are Scott's and Amundsen's expeditions to the South Pole. Let's score these as our first two examples.

Scott

1. Goal clearly defined? Not really. Only after he'd set out for the Pole did it actually become an objective. Half marks. (10)

2. A list of jobs so that everybody knew their part. No. Hardly at all. People's roles were confused and kept changing. (6)

3. One leader. No problem here. (10)

4. Jobs assigned to people. As mentioned earlier, there was a problem with people's roles and what was expected of them. (4)

5. Fallback position. History proves there wasn't. ( “15)

6. An appropriate leadership style? Scott had the one (autocratic) style that he used with everybody. I would score it low. (4)

7. Know what's going on. Given that most of the plan was in Scott's head he had to have a reasonable knowledge of what was going on. Where he would have fallen down would have been in understanding undercurrents and morale -type issues. (6)

8. Tell people what's going on? Hardly at all. (2)

This gives a profile as shown in Table A.1:

Table A.1.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	10
2	20	6
3	10	10
4	10	4
5	10	“15	15
6	10	4
7	10	6
8	10	2
9
10			12
		Grand total	27

Amundsen

1. Goal clearly defined? Absolutely. (20)

2. A list of jobs so that everybody knew their part. Completely. But let's say he could have done a little better. (18)

3. One leader. No problem here. (10)

4. Jobs assigned to people. Yes. See Chapter 4 (or better still read Huntford (1993) if you doubt it). (9)

5. Fallback position. Yes “ they put on weight, remember? (10)

6. An appropriate leadership style? Yes, Amundsen was very sensitive to his people. (9)

7. Know what's going on. Absolutely. (8)

8. Tell people what's going on? Yes. (8)

This gives a profile as shown in the Table A.2:

Table A.2.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	20
2	20	18
3	10	10
4	10	9
5	10	10	67
6	10	9
7	10	8
8	10	8
9
10			25
		Grand total	92

Project X1

This is a project I worked on about ten years ago. It was the one I mentioned in Chapter 3 (Case Study 4) where there were two leaders.

1. Goal clearly defined? Sort of. There were no real specifications, but there were people on the project who had a good knowledge of what the system was trying to deliver. Give it half marks, roughly . (9)

2. A list of jobs so that everybody knew their part. We knew what we had to do from day to day; but did we know how that fitted into the bigger picture, and whether we were making progress? Definitely not. (6)

3. One leader. No, there were two. (1)

4. Jobs assigned to people. People were certainly assigned to jobs. However, given that the job list may have been incomplete (see above) then we'd have to mark this low. (4)

5. Fallback position. Well, I suppose there was because when the excrement hit the ventilating device, management was able to reconfigure the project to deliver decreased functionality. (4)

6. An appropriate leadership style? There were two leaders, one autocratic and opinionated, the other wishy-washy and a bit of a moan. (4)

7. Know what's going on. See the comments above on job list and assignment of people. (4)

8. Tell people what's going on? Somewhat. (4)

This gives a profile as shown in Table A.3:

Table A.3.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	9
2	20	6
3	10	1
4	10	4
5	10	4	24
6	10	4
7	10	4
8	10	4
9
10			12
		Grand total	36

The Battle of the Somme

Here's one we mentioned earlier in the text, the opening of the Battle of the Somme in the World War One.

1. Goal clearly defined? Yes, no doubt about it. (20)

2. A list of jobs so that everybody knew their part. Yes, dock 'em a couple of points on the basis that they might have done slightly better. (18)

3. One leader. Yes. (10)

4. Jobs assigned to people. Yes, done very well. (8)

5. Fallback position. Nope. ( “15)

6. An appropriate leadership style? Leadership in the World War One has been the subject of much controversy and literature. The prevailing style in the military of the time was "don't think “ just do what you're told." I'll give them 5 on the basis that more responsibility and initiative should have been passed down the line. (5)

7. Know what's going on. The technology of the time made communications on the battlefield very, very difficult. Half marks. (5)

8. Tell people what's going on? See earlier comments on "do what you're told." Half marks. (5)

This gives a profile as shown in Table A.4:

Table A.4.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	20
2	20	18
3	10	10
4	10	8
5	10	“15	41
6	10	5
7	10	5
8	10	5
9			15
10
		Grand total	56

Project X2

This is another project I once worked on.

1. Goal clearly defined? Yes, pretty well. We had a detailed spec. and descriptions of the functionality that would be in each release. (16)

2. A list of jobs so that everybody knew their part. No, we never really managed to stabilize our lists. We would get lists written, and then a short while later we would be told that "everything has changed," and the lists were no longer valid. (10)

3. One leader. No. We had horrifying problems “ too numerous to mention “ with leaders. Let me score it 1 on the basis that we were in the 0 or more than 2 category. (1)

4. Jobs assigned to people. Well, because the job lists were in flux, so too were the job assignments. (3)

5. Fallback position. Yes, we did. As the going got tough we were able to move functionality from the first release into subsequent ones, a standard trick among software developers. (7)

6. An appropriate leadership style? Because of the problems described earlier, I eventually had to take control of this project and drive it through to a conclusion. Let me (modestly!) score myself 7. (7)

7. Know what's going on. By and large, yes we did. (6)

8. Tell people what's going on? We tried. (6)

This gives a profile as shown in Table A.5:

Table A.5.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	9
1	20	16
2	20	10
3	10	1
4	10	3
5	10	7	37
6	10	7
7	10	6
8	10	6
9
10			19
		Grand total	56

Project X3

Here's another software project I worked on. The project manager is one of the best I know.

1. Goal clearly defined? Yes. (18)

2. A list of jobs so that everybody knew their part. Yes. (18)

3. One leader. Yes. (10)

4. People assigned to jobs. Yes. (8)

5. Fallback position. Yes, even though it wasn't needed. (8)

6. An appropriate leadership style? Yes. (8)

7. Know what's going on. Yes. (8)

8. Tell people what's going on? Yes. (8)

This gives a profile as shown in Table A.6:

Table A.6.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	18
2	20	18
3	10	10
4	10	8
5	10	8	62
6	10	8
7	10	8
8	10	8
9
10			24
		Grand total	86

Operation Desert Storm

Clearly, I wasn't in on the planning of Operation Desert Storm. However, here's a man in the street's assessment of it.

1. Goal clearly defined? Yes, with clinical accuracy. (20)

2. A list of jobs so that everybody knew their part. Seems like. (18)

3. One leader. Yep, Mr Schwarzkopf. (10)

4. Jobs assigned to people. Seems to have been done well. (7)

5. Fallback position. My guess is that there was “ can't know for sure, though. (7)

6. An appropriate leadership style? Don't seem to come much better. (9)

7. Know what's going on. We didn't, but presumably the powers that be did. (8)

8. Tell people what's going on? Again, we didn't, but I think those involved did. (8)

This gives a profile as shown in Table A.7:

Table A.7.

STEP	POSSIBLE	ACTUAL	TOTALS
1	20	20
2	20	18
3	10	10
4	10	7
5	10	7	62
6	10	9
7	10	8
8	10	8
9
10			25
		Grand total	87

Analysis

Trends

The projects I've just described are shown in tabular form in Figure A3.1. From the limited sample presented here and/or studied by us, a few things have become apparent.

• Any project with a planning phase score below about 40 looks dodgy.

• Any project with a total score below 60 looks dodgy.

• If you get a low score from the planning phase, then putting it charitably, you have limited scope to improve things subsequently. Putting it less charitably, if you come out of the planning phase with a low score, then it would appear that you can do what you like and it won't really make a lot of difference to the eventual outcome.

This last observation gives rise to my corollary to the famous Brook's Law (Brooks, 1975) which states "Adding more people to a late project makes it later." My corollary states "Projects succeed or fail in the first 10 percent of their lifetimes."

Usage

What we have described here is a very simple risk analysis. It can be done at any point in a project's life. It is very much a garbage-in, garbage-out exercise; if you tell it lies, it will tell you lies in return.

Limitations

At the moment the limitations of "taking the PSI challenge" (if I may call it that) include the following:

• It is based on a very small sample

• The scoring is somewhat coarse

• The scoring is somewhat subjective

• It has had limited application in real life

• Both the scale and scoring process need to be tuned and refined by reference to many other projects

That much having been said, we are now using the PSI on consultancy assignments and on training courses. Any projects we have studied to date have verified the accuracy of our PSI scoring scheme. We're looking forward to doing more work in this area.

Figure A.9. PSI analysis table

Comments

• Scott. The worst project we have; ground zero on our scale

• X1. Scores very poorly as you would expect

• The Somme. Illustrates that in certain circumstances a margin for error can be a dealbreaker

• X2. Poor profile after planning just about saved by a strong implementation

• X3. Profile similar to Amundsen; strong plan, strong implementation

• Desert Storm. Almost as good as Amundsen, the best project we have

Rules of thumb

Anything coming out of the planning stage below 40 is very dicey. Failures will probably end up with scores of below 60.