Aptage Learning in a Small MSProject Example

Several things to note:

  • This is a trivial 2 task project. This is intentionally tiny to highlight the learning and insights provided by Aptage.
  • MSProject itself predicts % complete, and we don’t believe their calculation has any basis in reality.

Day 0: One Team, Two Tasks

This project is 2 tasks, each estimated at 5 days long, with a dependency that Second can start after First is finished. The dependency adds transitive risk (due to conjoint probabilities) and the initial confidence of completing both tasks at the end of 10 working days is 31%. It is assumed that a 5 days task may optimistically finish in 3 days, or may run long up to 8 days (this is our default “natural uncertainty”). If all goes well this project starts on April 1st and finishes by April 14th.

Project – One Team – Two Tasks.mpp

  • Today is April 1st, 2020 and we are at Day 0 of the project. 
  • Simple project, two tasks (First and Second) with dependencies between them. 5 days long each.
  • The 80% confident completion date is 2.41 weeks long.

Figure 1

Overall risk is 69%. This is higher than our natural uncertainty risk of 52% because of the dependency causing transitive risk on the critical path (due to conjoint probabilities).

Day 0: Two Teams, Two Tasks

The is the same project plan, but each task is assigned to a different team/resource. The implication of this is that good/bad news about one team doesn’t have any affect on the other team/task (and Aptage doesn’t learn from it).

Project – Two Teams – Two Tasks.mpp

The analysis on April 1st is the same for this project. At the start, we know exactly the same information for each team and task.

Figure 2

Day 7 – First Task Goes Long

The First task takes 7, not 5, days to complete. This is less than our worst case estimate of 8 days, but still not great news. There are now only 3 days remaining before the original deadline… and the Second task was originally estimated at 5 days.

Project – One Team – Two Tasks – First is Late.mpp

The confidence of an on-time finish are now negligible. In fact, the 80% confident completion time is now up to 3.24 weeks total project time. This project is forecast to finish between 2-4 week total execution time.

Figure 3

There are two sources of increased risk at this point in the project:

  1. Late start, we are trying to get a 5 day task done in 3 days.
  2. FirstTeam making slower progress (either estimated low or worked slower) than expected.

The higher risk due to the team getting work done slower than originally planned is visible in a velocity chart. For gantt or MSProject based plans we expect teams to be around 1.0 (that is do about 100% of what they expect to accomplish). The evidence shows the team is likely to be only at 60-80% original planned capacity.

Figure 5

Notice MSProject calculated 58% completed…

We Aptians don’t think 58% reflects a realistic assessment of completion. Using the 80% confidence mark of 3.24 weeks, and the 7 days of project work days put in so far we could calculate this to be perhaps 43% (7 days or 1.4 weeks divided by 3.24 weeks likely).

Day 7 – First Task Goes Long, Two Teams

The First task takes 7 days. This has no meaning for the SecondTeam and the Second task…

Project – Two Teams – Two Tasks – First is Late.mpp

The confidence of an on-time finish are still negligible, but with two teams the 80% confident completion date are very different. At 2.67 weeks this project is very likely to finish before the single team project.

Figure 4

There is only a single source of risk, the late start for the Second task. The velocity of the FirstTeam is still slowed considerably, but the SecondTeam (of which we have no learning or evidence about) is still at the default near 1.0 and this project is forecast to finish between 2-3 weeks of total execution time.

Figure 6

Why does the Figure 3 have some green near 2 weeks and this project doesn’t?

The answer is uncertainty. The wider the likelihood distribution the more uncertainty… but it goes both ways towards more late and towards more early.

About the author
Leave Comment

Your email address will not be published. Required fields are marked *

clear formSubmit