ADJUSTING YOUR PROJECT BUDGET FOR RISK
by Wayne Strider

We are all familiar with the reputation of software projects in general.
They take longer than planned and overrun their budgets by factors of
1.5 or greater, blah, blah.  Even when initial project estimates are
reasonably accurate, failure to adjust project schedules and budgets for
risk almost guarantees schedule slippage and budget overruns.

One way to manage risk is to add a lump sum contingency to the
budget such as 10% of the total project budget estimate. This
acknowledges the reality of projects. We know something will go wrong,
we just don't what or when or what the impact will be.  This is better
than ignoring the reality, however, there is more that can be done.

With a little work (okay, maybe more than a little) project risks can be
identified and analyzed for probability of occurence and severity of
impact if they do occur.  You may choose to mitigate the ones you think
can be prevented or contained, i.e., impact reduced.  Mitigation costs
money so you'll want to adjust the budget accordingly.  Other risks you
may decide not to mitigate because the cost is prohibitively high and
the probability or severity is low. But if those risks do occur you'll
want to plan how you will handle them and estimate the additional time
and cost and adjust the budget for this contingency.  One important
difference between mitigation and contingency is that you only spend the
contingency budget if a risk actually occurs.  Mitigation budget is
spent regardless of whether a risk occurs or not in order to prevent or
contain the risk.

Here is a simplified example to illustrate some of the choices
involved in adjusting your budget for risk.  I tried to keep the example
simple by excluding the math because I don't want you to get distracted
by the math.  If you have questions about the math please contact me at
816 746 8100 or waynestrider@striderandcline.com.

Overview of Project Background and Assumptions:

The Plan
Your project staffing budget is based on using 10 skilled in-house
developers with knowledge of the business.  Initial project staffing
estimate is $936,000.  Project duration is 12 months for the entire
project.

The Risk Identified is Staff Turnover
Your organization historically has experienced a 30% turnover rate per
year for developers.  You risk loosing three developers during your 12
month project.  In this example assume you will loose three developers
after 2 months with 10 months (43 weeks) left to go on the project.

Contingency Plan
You will have to replace lost developers as quickly as possible,
hopefully within the 2 weeks notice they give. Assume you have to pay
$90/hr. premium rate for contract developers on very short notice.
Contract developers will each require 2 weeks of familiarization with
the business and 3 weeks to become fully productive on the development
work for a total of 5 weeks. The total weeks for each developer will be
5 weeks of getting up-to-speed and 43 weeks of work for a total of 48
weeks. The familiarization work will require 2 weeks of a business
analyst's time and 3 weeks of your lead developer's time at $45/hr.,
significantly reducing their productivity.

Mitigation plan
You could decide to add up to three extra in-house developers at $45/hr
to the project from the beginning, so that the project budget and
schedule will not impacted when you loose three developers.  The extra
developers could learn the application as the project progresses along
with all the developers.  For this example, assume you will add two
developers to mitigate this risk.  Assume you will pre-arrange for one
developer on demand at a lower rate of $75/hr. with a contract vendor.
This will reduce the contingency cost to only the cost of one
replacement developer and the extra schedule time required to only 5
weeks. You could get lucky and lose only one developer in which case you
budgeted more than you needed.  However, with the extra resource you may
be able to finish the project sooner.

Schedule Adjustment
Adjust the project schedule by 5 weeks each time you loose a developer,
especially if the developer's work is on the critical path. The
schedule impact due to 3 lost developers would be 15 weeks slippage.  By
using the mitigation plan, you could reduce this schedule slippage to 5
weeks.

Here are four choices available to you.  What does your wisdom tell you?

1. DON'T EVEN THINK ABOUT RISK.  This looks like the least expensive
choice.  But is it really?  What happens when you get hit with replacing
3 developers?  There is no money in the budget to replace them.  Your
project is unprepared.



Initial Staffing Budget:
			
$936,000



2. ADJUST FOR CONTINGENCY (spend only when 3 developers leave)



Initial Staffing Budget:
			
$936,000


Estimated Contingency Budget:
			
$313,200


Estimated Staffing Budget
Adjusted for Contingency 133%:
			
$1,249,200



3. ADJUST FOR MITIGATION (add two developers from the beginning)


Initial Staffing Budget:
$936,000


Estimated Mitigation Budget:
$187,200


Estimated Staffing Budget
Adjusted for Mitigation 120%:
$1,123,200


4. ADJUST FOR REDUCED CONTINGENCY (one pre-arranged vendor developer)
AND MITIGATION (two in-house developers)


Initial Staffing Budget:
$936,000


Estimated Mitigation Budget:
$187,200


Estimated Contingency for one
developer only at $75/hr.:
$75,600


Estimated Staffing Budget
Adjusted for reduced Contingency
and Mitigation 128%:
$1,198,800




Which choice you make will depend on this project's tolerance for
schedule slippage and budget overrun. In this example, if the project
absolutely cannot stand more than 5 weeks schedule slippage, then the
wise choice would be the combination of mitigation and contingency.

Each project situation will be different.  For each project
identify the choices, do the math, and get other key project
decision-makers to help you choose the risk management approach that
makes sense for that project.

Strider & Cline, Inc. Wisdom Box Newsletter Vol. 2003, No. 1