Schedule Logic Check Metrics

Schedule Logic Check Metrics

by | 12 May, 2022 | Blog

One of the fundamental questions for any schedule health assessment is “Is the schedule logically driven?” There are some scheduling metrics that could help planners and schedulers to find the answer to this question. In this post, we will review some of them, that are related to missing dependencies.  

Missing Dependencies Metrics

Missing logic metrics is a group of metrics known as:

 

  • missing dependencies,
  • missing predecessors and successors,
  • open start & finish,
  • dangling activities.

This check is not as simple as many project managers and schedulers might think. In fact, it is actually very hard to answer the question: Are there any missing or incorrect dependencies in the schedule?

A very popular DCMA-14 Points schedule assessment includes a check of logic (metric N1). Schedulers who are familiar with a scheduling tool (usually Primavera or Microsoft Project) but lacking scheduling knowledge very often apply DCMA-14 “blindly”. Their schedule may have critical logical issues but is reported as a schedule with sufficient quality. I have seen this issue in small business projects and also in large construction programs. So, let’s review what needs to be considered for a comprehensive schedule logic analysis.

Missing Dependencies?

There is no metric that could confirm that a schedule has no dependencies missing.

The fact that each activity (except first and last) has a predecessor and successor doesn’t mean that another successor from/to this activity is not missing. The only way to guarantee that a schedule has correct logic is to implement correct schedule development and maintenance processes and apply control to ensure these processes are followed.

Each reporting period logic changes have to be analysed, documented and explained.

While scheduling metrics can’t guarantee that dependencies are not missing, some of them are good indicators that logic has to be checked and, if required, fixed. There are four primary schedule quality metrics used to indicate that a schedule may have a missing dependency:

♣   Missing Predecessor

All ‘Not Completed’ activities except 1st activity and external incoming activities must have a predecessor(s).

♣   Missing Successor

All ‘In Progress’ or ‘Not Started’ activities except the last activity and external outcoming activities must-have a successor(s).

♣   Open Start

Activities where only the predecessor(s) are either ‘Finish-to-Finish’ or Start-to-Finish resulting in an open start to the activity. All ‘Not Completed’ activities except 1st activity and external incoming activities must have at least one ‘Finish-to Start’ or ‘Start-to-Start’ predecessor.

♣   Open Finish

Activities where the only successor(s) are either ‘Start-to-Finish’ or ‘Start-to-Start’ resulting in an open finish to the activity. All ‘In Progress’ or ‘Not Started’ activities except the last activity and external outcoming activities must have at least one ‘Finish-to-Start’ or ‘Finish-to-Finish’ successor.

There is a number of points that have to be taken into consideration when these metrics are applied:

  • A project may have external dependencies and it is not always possible, and in some cases, not recommended to link schedules from different projects. Milestones representing such dependencies do not have predecessors or successors.
  • A Project may have Level of Effort (LoE), Hammock and WBS activities. Different scheduling tools implement these types of activities differently and this impacts the “missing logic” analysis. 

–   Hummocks in MS Project are shown without a predecessor and a successor but there is no way to identify “Hammock” type of activities in the system.

–   “WBS summary” activities in Primavera don’t need a predecessor or a successor. So they have to be excluded from the analysis. Another metric is required to check that “WBS summary” activities don’t have logic as Primavera permits linking activity to “WBS summary” activity.

–  In Primavera, if all activity successors are linked to LoE or “WBS summary” activities, the activity is actually missing a valid successor. Otherwise, based on the logic, there are no requirements to complete this activity. The same is applicable to predecessors. Such cases could only be identified via a specially developed project report, not via filters.

  • If a milestone has “Open Start” or “Open Finish” it actually doesn’t create any issues with logic. Milestones could be excluded from these metrics.
  • Microsoft Project has a unique feature to link activities with summary tasks. If this technique is applied it is very hard to analyse logic, as some of the activities are driven by logic from activities and some from summary tasks, so it is not recommended. An additional metric is required to identify summary tasks with processors and successors.

Filters

Missing logic metrics could be developed via filters in Primavera and Microsoft Project but, as described above, specifics of each tool have to be taken into account. Spider Project already has all these metrics build-in and also allows developing comprehensive additional filters as required.

Primavera filters to identify activities with missing logic:

♣  No Predecessors

♣  No Successors

♣  Open Finish

♣  Open Start

External Schedule Analysis Tools

When an external schedule analysis tool is used each metric has to be configured to address explained challenges. For example, Acumen Fuse has a pre-developed “Missing predecessors” metric. However, the metric includes ALL activities with missing predecessors. If an activity already started (or completed) it doesn’t matter if it has a predecessor or not. It is recommended to configure this metric to exclude these activities. Otherwise, the Acumen Fuse Report creates “noise” and may incorrectly show that the schedule has predecessors issues when it actually doesn’t.

Alex Lyaschenko

PMO | Portfolio Planning & Delivery | PMP | P3O Practitioner | AgilePM Practitioner | Six Sigma

Critical Path Drag

Critical Path Drag

by | 4 May, 2022 | Blog

Majority of project managers are familiar with the critical path method (CPM) and think that they know how to apply it to manage their projects.

Many scheduling software can compute and visualise Critical Path, so it is a no-brainer to use it. Usually, a project manager accepts the calculation as the schedule. The Critical path method doesn’t take resources, material and financial supply constraints into the consideration. So, the next steps become to assign resources, baseline the schedule and start to manage project progress.

This is because the project managers have forgotten that the “M” in CPM stands for method. The reality is that the first network is only the tentative schedule, giving just the data to implement the method fully. Once the initial CPM schedule has been computed and a logic diagram produced, the project manager is in a position to use the critical path method to save time and money. If, for instance, we want to shorten the project duration, we now know where to start: the current critical path!

There are three ways of trying to shorten the project:

1. Shorten the duration of activity on the critical path by:

  • adding extra resources
  • assigning resources with better productivity
  • ask resources to work more hours
  • change delivery method (a backhoe instead of 10 labours)
  • prune scope (gold planning)

2. Start critical activities earlier by:

  • applying anti-crashing method

3. Change logic and remove activity from the critical path by:

  • applying fast-tracking method

But the question is: Where should we add/remove these resources? Where should we review the scope? What will the effect be of such actions on the schedule and cost?

For the answers, we need CPM metrics.

One of the CPM metrics that could help us find the answers to the above questions was proposed by Stephan A. Devaux and explained in his “Total Project Control” book called Critical Path Drag. In the 1st edition, published in 1999, Stephan used the term DRAG (Devaux’s Removed Activity Guide), but since the 2nd edition, it was written as “Drag”.

In his book, Stephan described the Critical Path Drag metric as:

Critical Path Drag

Critical Path Drag is the quantification of the amount of time each activity is adding to the project.

To some extent this metric could be considered as an opposite to the Total Float metric.

Total Float

Total Float is the amount of time an activity can be delayed before its path becomes the longest path.

By contrast, drag is:

  • only#1 on the critical path; and
  • the amount of time by which an activity duration can be shortened#2 before its drag is reduced to zero.

Further decrease in the duration of the activity would not reduce the duration of the project as another path became critical. In other words, it is the amount of time that could potentially be saved on the project by reducing the duration of the activity (or removing an activity completely).

#1,2 The proposed explanation covers the majority of project scenarios, but not all of them. We will review such scenarios below.

Most of the software that computes Critical Path also could calculate Total Float (TF), but just a few can calculate Critical Path Drag, so this metric is still not well known. However, this computation is very important. Without understanding activity drags, well too often, implemented project optimisation measures don’t give the expected effect.

The formula for computing drag on a simple critical path network schedule is as follows:

  • If an activity is off the critical path, its drag = 0
  • If an activity is on the critical path and has nothing else in parallel#3, its drag = its duration
  • If an activity is on the critical path and has other activities in parallel, its drag is the lowest number between:

– activity duration
– total float of the parallel activity with the least total float

A programming formula may look something like this:
= IF (Critical = “Yes”, Min (duration, min({total float of parallel activities}), 0)

#3 Parallel activities belong to a parallel stream. They are not necessarily planned to be performed at the same time as the analysed activity.

Critical Path Drag Calculation

Let’s review the following simple project:

With all “Finish-to-Start” dependencies, the duration of this project is 18 days. Drag for first and last activities is equal to the duration of activities.
The drag for the second critical activity is only 3 days. This means that if the activity will be reduced by 3 days, the project also will be reduced by 3 days. Any further optimisation would not decrease the duration of the project as this activity would not be on the new critical path anymore.

Let’s assume it is possible to add resources to the 10-days activity and complete this activity twice quicker. The new duration of this activity will be 5 days:

The project has a new Critical Path and as we expected duration decreased to 15 days.

Drag for 2nd and 3rd critical activities is 2 days each. However, it doesn’t mean that if both activities are reduced by 2 days, the project will be 4 days shorter! The drag is not cumulative.

Negative Critical Path Drag

Let’s consider a slightly more complex example: a working typical fragment “1 km of Road Construction”. There are only 20 activities in this fragment, but there are much harder to calculate critical path drag manually. Let’s use a scheduling tool, Spider Project, to calculate CP drag for each activity:

1 km of Road Construction.

There are 10 critical activities and all of them have Critical path drag <>0. In this project fragment, the first and the last critical activities have CP drag equal to the duration of the activity. In other cases, drag is less than the activity duration. Also, there are 3 critical activities where drag is less than zero. Let’s analyse how it is possible.
As it was explained in the “Project Anti-Crashing Method” post, in some cases project duration could be reduced by increasing a critical activity duration:

In this example, the CP drag of the activity “C” will be negative. If the duration of this activity is reduced to zero, the duration of the project will be increased from 42 to 50 days. So, CP drag will be minus 8 days.
At the same time, if it is possible to increase the duration of activity “C” by 14 days, the duration of the project will be decreased from 42 to 28 days (!)

Critical Path Drag on non-critical activities

It is logical to think that project duration could be reduced by optimising activities on a critical path only. However, there are project scenarios when decreasing the duration of the non-critical activity could reduce the overall duration of the project.
In the previous post, we reviewed a project fragment when activities have different calendars.

Activity “A” only could be performed on weekdays and activity “B” only could be performed on weekends.

Activity “A” has 4 days of the total float, so it is not on the critical path. However, if we reduce the duration of this activity to 5 days, the overall duration also will be reduced from 14 days to 7 days.

So, the CP drag of activity “A” will be >0. How much exactly? Well, it depends..
If it is based on the activity “A” 5-days calendar, CP drag is 5 days. If 7-days calendar, CP drag is 7 days.

Summary

  • The Critical Path method gives project teams information to optimise their project plans, not the final schedule. The method has optimisation metrics for that.
  • Activity Drag is an important Critical Path metric that allows prediction of how the applied effort would shorten project duration
  • In some cases Activity Drag could be negative
  • When different calendars are applied non-critical activities may have a positive ‘Activity Drag’. Optimisation of these activities would shorten the project duration.

We are going to review other Critical Path optimisation metrics in future posts.

Alex Lyaschenko

PMO | Portfolio Planning & Delivery | PMP | P3O Practitioner | AgilePM Practitioner | Six Sigma

Hidden Project Schedule Opportunities

Hidden Project Schedule Opportunities

by | 3 Apr, 2022 | Blog

A few days ago, I had organised a poll on LinkedIn and asked a simple question: Can project duration potentially be reduced by decreasing the duration of non-critical activity (total float >0)?

While the poll is still open, and the final result may still slightly change, it is already clear that half of the responders believe that reducing an activity that is not on the critical path would never change the duration of the project, and another half don’t think so. Most of the project participants believe that they know how the critical path method works, so why is there no agreement on a simple question then?
By organising this poll, I wanted to demonstrate the fact that project management is not as simple as it is presented in project management trainings, and even in global standards or organisational frameworks (PMBOK, NASA, etc.). The courses and standards always simplify reality to make it easier to understand.
Apart from project management, it is hard to find another area where professionals continue to apply simplified methods to solve complex and complicated problems, and real project management is never simple. To pass a project management exam, what you need to know is not the same as what we need to know to manage projects!
A real-life project is not just a list of activities with ‘Finish-to-Start’ type of dependencies, as is shown in almost all books that explain project management methods (incl. critical path method) and techniques.

It also includes:

  • Start-to-Start, Finish-to-Finish and even Start-to-Finish dependencies;
  • Double dependencies;
  • Resource limits;
  • Hard & Soft constraints;
  • Different resource & activity calendars;
  • Lags & leads;
  • Risks & uncertainties.

Even the critical path method may not work as expected when these aspects are taken into account.

Let’s now go back to the question in the poll and find the correct answer. So far, in comments to the poll, there are no examples offered, that would demonstrate how a non-critical activity could potentially reduce the duration of the project. Let’s review a small project fragment:

Example

The fragment contains two linked (Finish-to-Start) activities with different calendars. The predecessor activity “A” has six days of duration and a 5-days (Mon-Fri) calendar. The successor activity has 2 days of duration and a weekend (Sat-Sun) calendar.

The duration of this fragment is 14 days. The activity “A” is not on the critical path as it has four days of the total float. However, if we could reduce this activity to 5 days, the overall duration will be reduced to 7 days:

This simple example demonstrates that in some cases, project duration could be reduced by decreasing the duration of a non-critical activity.
It is just one of the examples of “hidden opportunity” in our schedules. There are many others. Such opportunities often exist, but they are not obvious. Unfortunately, MS Project and Primavera don’t have built-in critical path metrics that could help project teams to identify project opportunities. I am going to explain such metrics in future posts.

Alex Lyaschenko

PMO | Portfolio Planning & Delivery | PMP | P3O Practitioner | AgilePM Practitioner | Six Sigma

Project Anti-Crashing Method

Project Anti-Crashing Method

by | 24 Mar, 2022 | Blog

PMBOK and some other project management standards describe two key methods for schedule optimisation: fast-tracking and crashing.

Apart from these well-known methods, there is a method that hasn’t been described widely. So, let’s feel in the gap.

Anti-crashing

This method works as it sounds: opposite to the crashing method. So, let’s review the project crashing method first.

Crashing technic in scheduling:

Crashing is a project management method used to speed up a project’s timeline by adding extra resources without changing the project’s scope.

Adding extra resources is not the only way to reduce the duration of an activity.

  • Activity duration can also be reduced by assigning resources with higher productivity rates or by changing resource calendars: extended working hours, work on weekends and holidays, additional shifts, etc.
  • Project duration can be reduced only when crashing is applied to activities on the critical project path.
  • Just like two women can’t deliver a baby in 4.5 months, the duration of some activities can not be reduced by deploying additional resources
  • Duration of activities does not always have a linear dependency on a number of assigned resources. This issue was discussed in the “Project Team assignment” post.
Project Team assignment

The crashing method sounds very logical and easy to understand. However, project life is not as simple as it’s described in project management books. In real-life situations, there are cases when project duration can be shortened by REDUCING assigned resources.

Let’s review the following example:

Project Critical Path: A – B – C – D – E

Project Duration: 42 days

Resources assignments:

Activity “B”: 2 resources

Activity “C”: 2 resources

Activity “D” 2 resources

A project manager wants to speed up this project and has an opportunity to deploy additional resources to activities B, C and D.

 Let’s analyse how this potential optimisation would work:

 Assume critical Activity “B” has two additional resources assigned. Then the activity will be completed in 10 days (instead of 20 days), and the whole project will also be delivered 10 days sooner, equalling 32 days.

The same result will be seen, if Activity “D”, has 4 assigned resources (instead of 2).

 Now let’s apply crashing to Activity “C”. If 2 additional resources are assigned, this activity will be completed in 4 days (instead of 8).

The critical path will be the same, and the new duration of the project will be 46 days.

Wait, now the project will take 4 days longer, not shorter (!!!).

 What if we do the opposite and reduce the number of assigned resources in Activity “C” from 2 to 1?

In this case, Activity “C” will be completed in 16 days (instead of 8), and the new duration of the project will now be 34 days.

 So, as we can see, when we reduce the number of assigned resources, the duration of the activity will increase. However, the duration of the whole project will decrease (!!!).

It is not just a theoretical example; projects often have parallel activities performed by different teams or crews. In this case, start-to-start and finish-to-finish (sometimes lags and leads) logic is applied. Different teams/crews may have different productivity rates, and often the less productive team assigns extra recourses to their critical activity. However, applied crashing may sometimes lead to the opposite result. In those cases, projects need to apply the “anti-crashing” method instead.

Summary

Both crashing and anti-crashing methods focus on changes in resource management.

If the crashing method increases assigned resources to reduce the duration of critical activity, the anti-crashing method, on the opposite, reduces assigned resources to bring the start date of critical activity earlier.

Anti-crashing is a project management method used to compress a schedule by reducing resources and without changing the project’s scope. This includes:

  • Reduce assigned resources
  • Assigned resources with a lower productivity rate
  • Change working calendars

Actual schedules are much more complex than the example in this post, and it is not always easy to identify when anti-crashing needs to be applied instead of crashing. In the next post, I am going to explain how to identify these cases.

Alex Lyaschenko

PMO | Portfolio Planning & Delivery | PMP | P3O Practitioner | AgilePM Practitioner | Six Sigma

Realistic Project Delivery KPI

Realistic Project Delivery KPI

by | 22 Feb, 2022 | Blog

How often have we been told that projects are late and over budget and we need to do something about it?

At the same time, it is an absolute lack of understanding that it could be a sign of project delivery maturity rather than a real issue! The real issue is when ALL projects are on time and budget.

In this post, we will discuss this challenge from different perspectives and try to answer the fundamental question: Do we want all our projects to be on time and budget, or not?

An ideal project delivery

Many portfolio managers, sponsors, and project managers want their projects always to meet committed dates and budgets. An ideal scenario is when ALL projects are delivered on time and within budget, as their performance is often evaluated based on these criteria.

However, what is ideal for individual performance is not necessarily an ideal scenario for a portfolio. Let’s go deeper to a project level to understand why.

Each project ALWAYS has uncertainties and risks. So, it is not possible to predict project duration and cost precisely. It always ranges. A range for project durations and a range for project costs. Something like this famous hat:

Boundaries of these ranges may be unclear for project stakeholders, but they always exist.

In reality, there are rare cases when a range goes to infinity, but it is an exception from the rule and probably another good topic to discuss. The absolute majority of all projects have a distribution of cost and time.

There are many possibilities of how time and cost targets could be combined.

Someone (let’s say a “Sponsor”) decides which of all combinations will define project targets. The committed delivery date defines the target for the duration, and the committed project budget defines the target for the cost.

The probability of meeting both targets is lower than meeting each of the targets.

For example, if a project has agreed on targets and there is 90% chance that the project will be on time and a 90% chance that the project will be on the budget, the probability that both times and budget achieved will be less than 90%.

Different targets that include different types of contingencies may exist, but to keep it simple, assume there is one key set of targets. Delivery beyond these targets is going to be considered a failure.

Some sponsors choose very aggressive targets with low probabilities to meet them; others prefer a conservative approach with a higher probability of success.

When a sponsor doesn’t clearly understand ranges and distribution of possibilities (or are misled by low-quality risk analysis reports), they may assign a target outside the range. Yes, some projects are planned to fail.

Achievable and Realistic

If all projects in a portfolio constantly meet targets, this could mean one of two things:

  • Project managers have magic wands;
  • It was too easy to hit the targets;

Why could not all projects be just managed well and delivered as expected? Because Black Swan events exist.

A Black Swan event:

A Black Swan is an unpredictable event that is beyond what is normally expected of a situation and has potentially severe consequences. Black Swan events are characterised by their extreme rarity, severe impact, and the widespread insistence they were obvious in hindsight.

A Black Swan event is unpredictable (Unknown Unknowns), and you won’t find a related risk in the project risk register. Also, there are predictable (Known Unknowns), unavoidable low-probability / high-impact risks that could significantly impact project delivery. If such risks materialised and the project still meets the targets, it means that this project had a sufficient contingency to cover such risk, which could be the real problem!

From the “Parkinson’s Law”, we know that work expands to fill the time available for its completion. So, projects with significant contingencies (to cover “Black Swan” events) will likely take longer and be more expensive, even if the risks don’t materialise.

Parkinson’s law:

work expands to fill the time available for its completion.

The project delivery paradox

Mature and non-mature portfolios have different approaches to setting up projects and controlling project delivery.

Organisations with a non-mature project delivery don’t plan projects properly or apply deterministic, critical-path-based planning. Executives assign project targets without visibility of possibilities. Their projects take longer and spend more (sometimes substantially more) budgets. However, Portfolio managers rarely understand the actual delivery rate and often have an illusion of good performance.

Once executives of one of my clients decided to increase the project delivery rate from 90% to 95%, their Project Delivery KPI report showed that they were constantly above the 90% threshold. In reality, the actual measures discovered that they were below 40%.

Organisations with the mature project and portfolio management apply probabilistic planning and set the targets based on calculated probabilities. They accurately monitor how the probabilities change during the project’s progress. Their project delivered faster and with less budget, but it is still below the 100% mark. The 80% of on-time & budget projects is usually a good result.

A project portfolio delivery methodology that includes probabilistic methods is SDPM (Success Driven Project Management). SDPM has methods for:

 

  • full integration of scope, time, cost and risks
  • development of probabilistic project delivery plan and schedule
  • set and monitor probabilities of success.

Organisations that implemented SDPM understand how changes in project delivery and a risk profile impact probabilities to meet agreed targets. They control the contingency consumption.

Summary

  • A project with a contingency that covers low-probability / high-impact risks and ‘Black Swan” events is likely to meet planned targets. However, the project is also likely to spend available contingency even if the risks have been avoided (Parkinson low).
  • When all projects in a portfolio always meet targets, it is a sign that project KPIs don’t reflect reality, and the genuine delivery rate could be much lower.
  • SDPM methods focus on building optimised probabilistic project delivery plans, setting up the achievable project and portfolio targets, and controlling probability changes.

Alex Lyaschenko

PMO | Portfolio Planning & Delivery | PMP | P3O Practitioner | AgilePM Practitioner | Six Sigma