Project Management, as Taught by Eli Goldratt · Lesson 2 of 7
The project management cloud
Build the conflicts behind three complaints, late projects, changes and rework, and watch them generalise into one cloud every project manager lives in. Then the three ways to attack it, and the one that leads to a generic solution.
A cloud for each complaint
A conflict always has the same five-part structure: an objective, two needs that are each necessary for it, and two actions (requirements) that the needs seem to demand, which can’t both be done. Some questions help turn a complaint into its cloud:
- D: what action do you take to deal with the complaint?
- B: what need are you trying to satisfy by that action?
- C: what need is jeopardised by that action?
- D′: what do you do instead, to protect that need? (Usually: not take the action.)
- A: what objective do both needs serve?
Complaint 1: original due dates are not met. You’re in charge of a project about to miss its date. To bring the schedule back on track, you must take expensive corrective actions or trim the content. To keep the original commitment on content and budget, you must not. Both serve one objective: meet the original commitments.
Complaint 2: too many changes. A significant change is needed. To honour the commitment to deliver what the client really needs, you must make the change. To meet the original due date and budget, you must not. Objective: meet the original commitments.
Complaint 7: too much rework. Part of the project must be redone. To deliver the desired content, you must do the rework. To stay within the original budget and time, you must not. Objective: meet the original commitments.
Generalising
Put the three clouds side by side and look for the common aspect of each box:
| Box | Cloud 1 | Cloud 2 | Cloud 7 | Generic |
|---|---|---|---|---|
| D | Take expensive corrective action or trim content | Make the change | Do rework | Compensate for early mis-estimations and mis-considerations |
| D′ | Don’t | Don’t | Don’t | Don’t compensate |
| B | Get the schedule back on track | Deliver what the client really needs | Deliver the desired content | Do whatever it takes to meet the endangered original commitment |
| C | Keep content and budget | Keep due date and budget | Keep budget and time | Don’t jeopardise any other original commitment |
| A | Meet original commitments | Meet original commitments | Meet original commitments | Meet the original commitments |
All the undesirable effects of managing a project are direct consequences of this cloud. (Managing a project shouldn’t be confused with the two steps before it: choosing the right project and determining its content. This session doesn’t address those.)
Three ways to attack the cloud
Compromising between D and D′ leads to the UDEs, so there’s no point continuing to compromise. The TOC way is to challenge the assumptions under the arrows. A good solution removes the conflict without creating significant drawbacks.
Attack A–B. Assumption: the original commitment is realistic. Meaning: accept that the original commitments aren’t realistic, and institute procedures that ensure only realistic commitments are given. Drawback: you lose everything that ambitious commitments bring, such as your competitive edge.
Attack C–D′. Assumption: meeting the endangered commitment will unavoidably jeopardise other commitments. Meaning: find a way to meet it without hurting the others, a breakthrough, such as a technology that is cheaper and faster at the same time. Drawback: you can’t rely on finding a specific breakthrough every time. It isn’t a generic solution.
Attack B–D. Assumption: the safety we are allowed is not enough to absorb all the glitches. Meaning: there is a lot of safety embedded in task estimates, enough to absorb all the glitches, but the way we manage the project wastes it. Application: change the way we place and use the existing safety.
Pause and think: which attack can become a generic solution for project management?
- A–B: only give realistic commitments, padded enough to be safe every time
- C–D′: look for breakthroughs
- B–D: stop wasting the safety that is already in the estimates
It needs no luck and costs no competitive edge, provided the safety is really there and really wasted. A project manager’s first reaction: “I embed a lot of safety and waste it? Explain that step by step.” The next lesson does.
Check your understanding
Answer all five. Get four right to complete the lesson.
1. In the cloud for “too much rework”, the two actions are…
- Hire or fire
- Do the rework, or don’t
- Change scope or budget
- Delay or expedite
To deliver the content, or to stay within budget and time.
2. The generic action D is…
- Add resources
- Cut scope
- Renegotiate the due date as soon as a slip is detected
- Compensate for early mis-estimations and mis-considerations
Every complaint’s action is a form of compensating.
3. The drawback of attacking A–B is…
- Losing the benefits of ambitious commitments, like a competitive edge
- It needs new technology to deliver faster and cheaper at once
- It costs more safety
- None
Realistic but uncompetitive commitments.
4. Attacking C–D′ depends on…
- Better estimates from more experienced planners
- Fewer projects
- Finding a specific breakthrough each time
- Buffers
Not a generic solution.
5. Attacking B–D means…
- Adding more safety to each task, so that glitches are absorbed
- Changing how the existing safety is placed and used
- Removing all safety
- Blaming estimators
The safety is there; it is being wasted.