Welcome to Project Management!
In this video, I’ll talk about project crashing. Crashing is a term for shortening the duration
of an activity or project beyond its normal duration.
There could be many reasons for compressing project durations.
The time-to-market pressure is one of them. We mentioned this in a previous video: a 6-month
delay in bringing a product to market can result in a loss of market share of about
33% for a high-tech company. Another reason is unforeseen delays. For example,
5 months into the project, you realize that you are already 6 weeks
behind the schedule for the project. You need to do something to bring the project back
on schedule. A third reason is the high overhead cost.
It may cost $70,000 per day to just house and feed a construction crew in Alaska. You
would want to finish the project as quickly as possible.
Some project managers want to finish some project activities quickly because they are
under pressure of losing the equipment and personnel to other ongoing projects which
they are sharing resources with. Sometimes top management make a public commitment
for an early product delivery date without consulting the project manager or the team
members. The project manager has no choice but reducing project duration to meet the
deadline. Incentive contracts are also good reasons
for shortening the project duration so the project team can get significant bonuses for
early completion. Let’s see an example.
In 1994, a 6.8-magnitude earthquake struck the Northridge area of Los Angeles. It collapsed
sections of the highway system and disrupted the daily commute of 1 million people.
The California State Governor’s Office reported that for every day the freeway was closed,
it costed the local economy more than $1 million. To expedite the recovery process, the government
offered attractive incentives for completing the repair work ahead of schedule.
A construction company called C. C. Myers won the contract for rebuilding some of the
collapsed Interstate highway bridges. They finished the project in only 66 days, which
is 74 days ahead of schedule, and earned a $14.8 million bonus!
In order to finish the project early, they spent a lot of money on overtime, bonuses,
special equipment, and other premiums to keep the job rolling quickly.
They hired 134 ironworkers instead of the normal 15. Special lighting equipment was
set up so that work could be performed day and night. The sites were prepared and special
materials were used so that work could continue despite inclement weather that would normally
shut down construction. Project managers have several effective methods
for crashing specific project activities when resources are not constrained. Several of
them are summarized here. We can simply add more resources like people,
finance and equipment to the project to make it go faster.
However, be mindful that doubling the size of the workforce will not necessarily reduce
completion time by half unless minimal communication is needed between workers.
For projects that involve frequent communication like software development project, Brooks’
law may apply. The law states: adding manpower to a late software project makes it later.
Another option is to outsource some part or the whole project to subcontractors.
The subcontractor may have access to superior technology or expertise that will accelerate
the completion of the activity. For example, contracting for a backhoe can
accomplish in two hours what it can take a team of internal laborers two days to do.
The third option is to schedule overtime. The easiest way to add more labor is not to
add more people, but to schedule overtime. You can avoid the additional costs of training
as well as coordination and communication encountered when new people are added. Also,
there are fewer distractions when people work outside normal hours.
Hourly workers are typically paid 1.5 times for overtime and double for weekends and holidays.
Because the pay is much higher, sometimes people treat overtime as benefits and have
to compete for it. The high-pay overtime may attract criticism if an organization is running
public service business. Watch this video to see the overtime pay in the BART system.
Just remember that BART stands for the Bay Area Rapid Transit in San Francisco.
Sustained overtime work may incur intangible costs such as health, divorce, burnout, and
turnover. It may also lead to an overall decline in productivity when fatigue sets in. Watch
another video to learn more about these intangible costs.
The fourth option is to establish a core project team with full-time concentration and avoid
multi-tasking. We talked about organizational structure in
a previous video. We learned that dedicated teams tend to complete projects much faster.
The last option I want to mention is called “do it twice”. That means we finish the project
with a simple solution quickly and then we make it better.
For example, when a bridge is damaged and needs repair, we can build temporary pontoon
bridges before the repair on the main bridge is complete.
Another example is that we can release software program version 1 and later correct bugs and
add new features in version 2. When resources are constrained, there are
only limited ways of accelerating a project. The first one is to improve project team efficiency.
The project team may be able to improve productivity by implementing more efficient ways to do
their work. For some projects, this is possible; but for others, it’s easier said than done,
especially when employees have already been pushed to their limits.
One approach called “I bet you” seems to work well in improving team members’ work efficiency.
Let’s hear this story. Clark is a project manager and he just received
a priority assignment from top management. The engineering sketches that were originally
due tomorrow need to be done by 4pm today. So, he approaches Josh, the team member who
is responsible for the task. Josh’s initial response is, “That’s impossible!”
Clark understands the difficulty and tells Josh, “I’ll make a bet with you. If you
can finish the design by 4pm, you’ll get two of the company’s tickets to tomorrow night’s
Celtics–Knicks basketball game.” Josh accepts the challenge, works feverishly
to complete the assignment on time, and is able to take his young daughter to her first
professional basketball game. Many project managers use bets like this to
motivate extraordinary performance. The bets range from tickets to sports and
entertainment events to gift cards to amusement parks and high-class restaurants.
The bets have even greater significance if it also benefits family members or significant
others. Being able to take a son or daughter to a
professional basketball game allows that individual to “score points” at home through work.
The principle behind the story is called the expectancy theory. It works if the answers
to the following three questions are all affirmative. Question number 1: Can I do it? Or is it possible
to meet the challenge? Question number 2: Will I get it? Or can I trust the project
manager and will he or she deliver what’s promised? And question number 3: Is it worth
it? Or does the payoff have sufficient personal value to warrant the risk and extra effort?
If the answer to any of these three questions is a no, then this approach won’t work.
Also, one piece of practical advice when applying this approach: bets should be used sparingly
under special circumstances that require extraordinary effort. Otherwise others may become jealous.
When resources are limited, another way of crashing project duration is called fast-tracking.
It requires the project manager to rearrange the project network so that some critical
activities are done in parallel with existing resources.
One example takes it to extremes. Instead of building a house in the sequence like foundation,
then interior, exterior, and finally roof, these components can be built concurrently
and assembled together. There are even foldable houses that can be
premade and then shipped to the customer. Watch this video.
Another option is to reduce the project scope. That means certain tasks, deliverables, or
requirements will be reduced or even eliminated. Focus should be on changes in activities on
the critical path, otherwise the project duration won’t be shortened.
This method may come at a cost of reducing the value of the project and must be communicated
with other stakeholders like customers and top management to get their approval.
Reducing quality is always an option to shorten the duration of a project, but it’s rarely
acceptable or used. If quality is sacrificed, most likely you
will end up with an unhappy customer and you are putting your business at risk.
Watch this video clip from the show Silicon Valley.
Project crashing will affect project costs. The general nature of project costs is illustrated
in this project cost–duration graph. The total cost is a sum of the direct costs and
indirect costs. As we mentioned in a previous video, direct
costs are those that can be assigned directly to a specific work package or project activity.
Examples include labor, materials, and equipment costs.
Indirect costs are those that are not directly associated with any particular work packages,
but they are linked to the entire project. Examples include supervision, administration,
consulting, etc. Direct costs decrease as project duration
increases. Indirect costs increase as project duration increases.
So, in general, crashing a project and make it shorter will increase direct costs but
reduce the indirect costs. We would like to crash the project to the
optimal duration, so the total cost is minimized. Next we need to decide which activities to
shorten and how far to carry the shortening process.
The basic principle is to shorten the activities on the critical path with the smallest increase
in cost per unit of time. This is a cost-duration graph for an activity.
This graph only shows the direct cost associated with the activity because indirect costs are
considered separately at the project level. To simplify our calculation, we assume that
the direct cost-time relationship is linear for each activity. So this is a straight line,
and not a curve. The normal point marks the normal time and
low cost for completing the activity. In this case it’s 10 days and $4000.
The crash point marks the limit—the shortest time possible with the increased cost. In
this case it’s 4 days and $10,000. The slope of the line represents a constant
cost per unit of time. It’s equal to the rise along the vertical axis over the run along
the horizontal axis, or the cost delta divided by the time delta.
In this case, it’s $10,000 minus $4,000 divided by 10 days minus 4 days. The result means
that for each day of duration reduction, the direct cost will increase by $1000.
A steeper slope marked by the red line means it will cost more to shorten the activity
for each day, and a less steep slope marked by the blue line means it will cost less to
shorten the activity for each day. So when we say we need to shorten activities
on critical path with smallest increase in cost per unit of time, we mean the activities
with the least steep slope or smallest slope value.
Now, let’s see a complete example. This project has 7 activities, A through G.
For activity A, the normal duration is 2 days, the normal direct cost is $3,000, the crash
duration is 1 day, and the crash direct cost is $5,000. Based on the equation we introduced
earlier, we can calculate the slope and it’s $2,000 per day, which is shown in the last
column. The calculation of other activities can be
done in a similar way. Note that the total normal direct cost of
the project is equal to the sum of all the normal direct costs of these activities, and
it’s $31,000. We assume the project network is already created
and it’s shown here. For each activity, we also show the normal
duration, crash duration, and slope in the table on the right side for reference.
The critical path is A-D-F-G, and it’s highlighted in red. The duration of the project is 2 plus
10 plus 4 plus 5 equals 21 days. Currently, the total direct cost is $31,000.
We need to always follow the rule to shorten the activities on the critical path(s) with
the smallest slope first. For these 4 critical activities, G has the
smallest slope, which is 0. However, it cannot be shortened anymore because
its normal duration and crash duration are the same, 5 days.
We use a little cross to indicate that the activity cannot be reduced any further.
The second smallest slope is Activity A, $2,000. So, let’s reduce the duration of activity
A from 2 days to 1 day. Reducing activity A by 1 day cuts the project
duration from 21 days to 20 days, but it also increases the total direct cost by $2,000,
from $31,000 to $33,000. We put a little cross to indicate that activity
A cannot be reduced any further. Now, we need to check if A-D-F-G is still
the critical path. Turns out it still is. Compare D and F. Activity D costs $2,500 to
shorten and activity F costs $3,000. Activity D costs less.
So we reduce the duration of D from 10 days to 9 days. The duration of the project is
further reduced from 20 days to 19 days. The cost of the project is increased by $2,500
to $35,500. Now we have two critical paths: A-D-F-G and
A-C-F-G. In order to reduce the project duration, we
need to reduce C and D together, which costs $5,500, or just reduce F, which costs only
$3,000. So we reduce the duration of F. The project’s
duration is reduced from 19 days to 18 days, and the total cost is increased to $38,500.
This move has created a network with three critical paths. The third one is A-B-E-G.
All activities are critical now. To further reduce the project duration, we
must reduce the durations of both C and D, together with one of these two activities,
B and E. Because E is cheaper, so we decide to reduce C’s, D’s, and E’s durations.
Now the duration of the project is 17 days. The cost of the project is $47,000. We still
have three critical paths. However, none of the activities on the critical
path A-C-F-G can be further reduced. That means we already reached the limit for this
project network. Although some activities like B, D, and E
can still be reduced, their reduction will not result in a shorter project duration.
Let’s summarize the project duration and direct costs in this table. The duration is reduced
from 21 days to 17 days, and the direct cost is increased from $31,000 to $47,000.
We assume the original total indirect cost is $40,000, and it will be reduced by $5000
as the project duration is reduced per day. The total cost is equal to the direct cost
plus the indirect cost. So, when the project duration is 21 days, the total cost is $71,000,
when the duration is 20 days, the total cost is $68,000, etc.
Finally, we find out that the lowest total cost is achieved when the project’s duration
is 18 days, and the total cost is $63,500. We plot that relation in this cost-duration
graph. It’s more straightforward and we can quickly find the lowest point in the total
cost curve. Further reduce the project duration to less
than 18 days will increase the total cost. In this example I just showed you, we only
used one criterion to decide which activities to shorten, that is to shorten the activities
on the critical path(s) with the smallest slope.
In real-world applications, the choice of activities to crash and how far to go rely
on many other factors as well. The logic of the cost-duration graph suggests
managers should reduce the project duration to the lowest total cost point. However, compression
of a project with several critical paths increases the sensitivity, reduces scheduling flexibility,
and increases the risk of delaying the project. The practical use of such analysis will probably
suggest only a partial movement from the normal time toward the optimal time, so it doesn’t
push to the edge. Another factor that needs to be considered
is that crashing an early activity may be ineffective. This is because sometimes subsequent
activities are likely to be delayed and absorb the time gained from crashing the early activity.
Sometimes, the resources required to accelerate a cheaper activity may suddenly not be available.
If the least-cost method repeatedly signals a subgroup to accelerate progress, fatigue
and resentment may set in. On the other hand, if overtime pay is involved,
other team members may resent not having access to this benefit.
So far, we have been talking about project crashing in terms of time. It’s also possible
to do project crashing in terms of cost. Organizations are always looking for ways
to reduce cost and get things done cheaply. Cost overruns create unhappy customers and
can damage future business opportunities. One commonly used option for cutting costs
is to reduce project scope. For example, on over-budget movie projects, the director may
replace location shots with stock footage to cut costs.
Sometimes cost savings can be generated by extending the duration of a project. This
is just the opposite of what we have been doing, shorten the project time with a higher
cost. A smaller workforce, less-skilled labor, less expensive materials, and cheaper equipment.
We may also ask the customer to take more responsibilities. For example, we may identify
tasks that customers can do themselves. Homeowners frequently use this method to reduce costs
on home improvement projects. It’s advised to negotiate with the customer before the
project begins. And this is only limited to areas in which the customer has expertise
and the capability to pick up the tasks. For example, a homeowner may agree to paint
the room instead of paying the contractor to do it. But the owner may not be able to
do plumbing and electrical wiring, which require professional qualifications.
The project manager can hold a brainstorming session to solicit cost saving ideas from
team members or other stakeholders. Project managers should not underestimate the value
of simply asking if there is a cheaper, better way.
Outsourcing project activities or even the entire project is also an option for reducing
cost. Many American firms outsource work to firms operating in India where the salary
of a software engineer is one-third that of an American software engineer. Outsourcing
customer service call centers to overseas can lead to a saving of 60%.
Watch this funny parody news video about outsourcing. When the project is outsourced, you have less
control over how it’s completed. Therefore you will need to have clearly defined deliverables.
Just be aware that there are also downsides of outsourcing. Watch the second video about
this. In summary, we talked about different options
for reducing project duration and cost in this video.
This is Yong Wang. See you in the next video.