Here’s a great article I happened upon while doing research for a paper I’m writing.  “Lessons Learned: IT’s Biggest Project Failures”  In this article we are treated to stories of IT projects that “first make people laugh and then” (hopefully) “make them think.”  As a long time student of the failed software project, I was neither surprise nor disappointed with the projects relayed.  The projects noted failed for reasons such as:

• Failure to perform a should-cost analysis before selecting a supplier
• Failure to recognize an unhealthy project before it was too late
• Unrealistic and aggressive timelines
• Scope creep.

Barak Obama's 2010 U.S. Federal Budget proposal promises a "New Era of Responsibility", and in the introduction he says,  

"...we must begin the process of making the tough choices necessary to restore fiscal discipline, cut the deficit in half by the end of my first term in office, and put our Nation on sound fiscal footing."

Tough choices indeed. There in lies the greatest challenge.  With the best intentions, our government tries to do good things, but always starts more projects than it can afford.  And often the expected "value" of an initiative is never fully vetted before a project is launched. 

How can you decide among project alternatives and choose an optimal strategy if you have not determined the value of each?

How can you restore fiscal discipline and cut the budget deficit if you do not know the cost-benefit analyses your projects? 

Since a new era of government spending responsibility will hinge upon making tough decisions, they must be supported by the disciplines of 
business case analysis, analysis of alternatives, and portfolio management optimization to be successful.


 These disciplines have been areas of study for us at PRICE Systems for the past two years and our work was released this week in TruePlanning 2009.  There are three innovative new features in this release:

1. Value models - enabling analysts to estimate the value and benefits of project once it is delivered, and over its lifetime
2. Business Case Analysis - enabling analysts to prepare and score a case for the project including the cost/benefit analysis, payback period, investment rate of return, strategic benefits, and other evaluation criteria scoring.  The business case tells the story of the project.
3. Analysis of Alternatives and Portfolio Analysis - enabling decision-makers to view Risk/Reward charts of alternatives and planners to reconcile total project estimates to budgets.

In this post I will give you an overview of these features and I hope to dive deeper in future posts.  Consider the decision a government planner must make on whether or not to launch a Business Process Automation (BPA) project.  The figure below shows a TruePlanning Payback view for the BPA project.



The three panels in this view tell the story of the project and make its business case.  The product breakdown structure view below identifies the elements of the project that will cost money and identifies the various benefits expected from the project implementation.



Each element in the tree represents either a cost item or a benefit associated with the project.  The cost items and benefit items are grouped into separate folders so that they can be reviewed individually or summarized.  

The Payback chart below shows how costs and benefits are realized over time so that project cost/benefit cash flow can be analyzed.



Each bar quantifies the costs and benefits realized in each year, the line shows the cumulative cash flow over time.

The Metrics table shown below shows the most commonly used financial metrics used for business case analysis.



This alternative of the BPA has a payback period of just over five years and has an IRR of nearly thirty-nine percent.

However, there is a problem.  When we combine our BPA project with the other projects in our portfolio we identify an estimated budget overrun.  As shown in the project portfolio view below, the sum of the project estimates exceeds our agency budget over time.




Now we must decide what we cannot do or what project we must cancel to stay within budget.  We know the business case for the BPA project,  how does that compare to the other projects?  The chart below shows the risk/reward for the projects under consideration. 



Tough decision time!  If you only had budget for two projects, which would you choose?  Are you willing to say NO to the others? 

TruePlanning 2009 has been delivered just in time to empower government decision-makers with the right tools so that they can "begin the process of making the tough choices necessary to restore fiscal discipline, cut the deficit in half by the end of my first term in office, and put our Nation on sound fiscal footing".  TruePlanning cost management software estimates software development costs, IT project costs, and now, the value of those projects. The value of determining a project's true value is more important than ever. 

Today, change is in the air.  As I write this, Barack Obama is about to be sworn in as our new U.S. President and the space community, among others, should be braced for change.  A recent LA times article reported that of the 74 questions asked of NASA by the Obama transition team, over half were on basic spending issues, including cost overruns.

The Obama team and the NASA Administrator Michael Griffin clearly do not see eye-to-eye.  Monday, it was announced that Mr. Griffin will step down from the post.  Griffin characterizes himself as an engineer and states that NASA shouldn't be evaluated by how well it estimated the cost of projects.

"We start these things out, and we admit up front we don't completely know how to do them. That is what makes them interesting," Griffin said recently.

"If we are to judge the worth of our work by our ability to estimate, then that is a standard I am not ready to apply or to accept," Griffin said.

Others are more realistic about NASA's obligation to create credible expectations for spending taxpayer money.

"Our space program is running inefficiently, and without sufficient regard to cost performance," wrote Alan Stern, a former NASA associate administrator who has been mentioned as a possible replacement for Michael Griffin, the current NASA administrator.

In a recent op-ed piece in the New York Times, Stern called the cost overruns a "cancer" that has cost the agency's science program about $5 billion over five years.

Agency officials said they had improved financial controls -- including forcing managers to better estimate costs.

Considering NASA's great technical achievements, one must wonder, "Is accurate estimating harder than rocket science?" 

I don't think so.  I believe the clue to accurately estimating the cost, uncertainty and risk associated with large, complex space projects can be found in a recent Defense AT&L article by Col. Brian Shimel, USAF.  Col. Shimel states,

"We cannot relieve ourselves of the need to plan for the future just because the future is uncertain. For our plans to be reasonably accurate and reliable, it is prudent we base them on rational analysis and not on wishful thinking.", and...

"We must think clearly about uncertainty and risk, and we must fight the temptation to discount those factors when communicating the real conditions of our management situation. We don't get in trouble because of risk and uncertainty. We get in trouble for not admitting to ourselves - and those who rely on us - all of the risk and uncertainty that inherently exist in everything we plan to do."

Our TruePlanning estimating and cost management software gives managers metrics and benchmarks of previous projects, and methods to realistically quantify uncertainty and risks.  TruePlanning's reports and charts expose over-optimism and show managers the real conditions of the situation.  I routinely recommend that government agencies quantify all the risks and uncertainties of each project in their portfolio and budget to the sum of the 70% confidence levels of each project, yielding 80% confidence that the projects can meet taxpayer expectations for the budget.

President Barack Obama just finished his acceptance speech.  He called upon us all to increase our service to the country.  Government managers can help by stepping up to the estimating challenge.

Accurate estimating is not rocket science.  There are sophisticated cost estimating models to accurate estimate software development costs, hardware development costs, IT project costs, and operation and support costs - we just need to be realistic, responsible and courageous about presenting the results for our analysis. Contact us and we'll help you get started today.

This was a fun and gratifying week at PRICE Systems. In our Mt. Laurel, NJ headquarters we had our annual holiday gift wrapping of presents we donated for needy children and families in the area.  The PRICE team gathered in our classroom and the wrapping began with no pre-instruction or guidance - but much merriment.  What a study in production process and learning!  Wrapping that first present was awkward.  It had been a while since I wrapped last.  How to measure the paper?  Where to cut?  How to keep the cut straight?  Which way do I fold first? How to fold?  What to do if I have too much paper on the ends?  Where do I put the tape?  ...and so on.  Needless to say that first present took some time, I wasted a lot of paper and tape, and the final product had some wrapping defects.  After reflecting on the results, I looked around at others to see what they were doing, asked some questions, processed that information along with my own experiences, and set out on the second package.  I did much better.  I wrapped faster, I wasted less, and the final product was an improvement over the first.  With each new gift I got better and better.  I "learned" how to be more productive, reducing the effort and materials with each present.  By the eight present, I started to focus on how I could reduce materials.  How can I use less paper?  How can I use the least amount of tape and still keep the paper form coming off?  I was improving the overall process.

Productivity increases like mine occur with each and every production process and it is an essential part of cost and schedule estimating.  The method most commonly used to model this effect is called "learning curve", and we use it in all of our models. You can find hundreds of references and equations on the web, but generally the model uses a "learning curve parameter" (LC) to estimate the cost for a certain production quantity.  The LC is expressed as a percentage, and (1-LC) represents the amount the production cost is reduced every time the quantity doubles.  For instance, if I say an 80% learning curve applies to the process, then every time the quantity doubles the cost to produce is reduced by 20%.  So if the first unit cost $100, then the second would be $80, and the fourth would be $64, and the eighth would be $51.20.  For the same starting point, a 90% learning curve represents less learning and higher costs, while a 70% learning curve steeper learning and lower costs. 





If I had timed my wrapping and measured the materials used, I could have used simple math to determine the learning curve that applied. I suspect it was close to 70%.  Determining the learning curve for everything from paper clips to space ships is more complicated, but that is one of the things we do at PRICE.  We perform operations research and analyze data to keep our models as accurate as possible, and embed that research in our cost management software.  Accurate learning curve estimates help our clients optimize their investments, prepare better budgets, plan more effectively.  And as part of this most recent study, we made  a lot of kids happier for the holidays.  Happy Holidays to you. 

Technology readiness is a critical cost driver of development programs.  Many high technology programs fail because initial cost and schedule expectations were based on the assumption that the technologies employed were proven,  when actually they were not. Space programs have the most dubious history in this regard.  I once listened to a Lockheed Martin executive explain how the X33 space shuttle was a great vehicle, but was canceled because it was, "two inventions short of meeting the requirements". Canceled after over one billion dollars were spent.




Starting development projects that have constrained budgets and schedule with unproven technology is a bad idea, one that has not gone unnoticed.  The Defense Acquisition Performance Assessment 2006 (DAPA) report states,

"Technology maturity or “knowledge-based” development has been a subject of considerable discussion between the Department and the Congress. However, although there is agreement concerning the advantages of ensuring that technology is mature prior to proceeding to development and production, there are no clearly definable measures of technology readiness.", and,

"Incorporating high-risk technology in systems generally leads to significant cost and schedule impacts."

So, have decision-makers learned their lesson by now. Recent news suggests not.  It was just published that NASA's Mars Rover mission was delayed two years and is over budget because,

"problems developed in the design and operation of the 31 actuators that control the mechanics of the craft, including the steering mechanism and its robotic arm"

Probably a few inventions short.

Technology should be matured in a lab environment with a sustained level of effort before it is employed in a development project with cost and schedule constraints.  Anything else is extremely risky and will most likely fail.

The PRICE TruePlanning models each have a method evaluating the cost and risks associated with technology maturity.  Our cost management software helps our clients decide among  alternatives with varying technology maturity and let them know the risk of overrunning cost and schedule.  Most of the time our analysis can dissuade executives from starting a project "two inventions short".

Jacob has a great comment to my recent post on getting it "right the first time".  He notes that requirements are often fuzzy and that estimates rely on peoples' opinion of cost and duration - and that the are often wrong.  He asked what silver bullet we have. Our silver bullet is a proven discipline that makes people better estimators and sheds light on fuzzy requirements.

It starts with the people. Becoming an accurate estimator requires the same step-by-step approach as any learned skill such as golf or tennis or swimming (things I am currently struggling to do well). First you need the right tools, then you need to learn the right way, then you need to be mentored by experts as you practice what you learned until it becomes second nature. Recently, I was struggling with my breathing while attempting to swim. Lap-after-lap I was breathing hard, swallowing water - it was a miserable sight.  Then I consulted an expert that got my head turning the right way and got me into an rhythm of breathing every third stroke.  My swimming quickly improved by 40% - more laps in less time.

We do the same for our clients.  First we arm them with the right tools.  Our TruePlanning software provides a proven step-by-step discipline to do an estimate the right way,  TruePlanning asks the right questions and covers all the bases, including risks.  Then we train people not just how to use our software but how to be great estimators.  Sizing methods, interview techniques, and requirements elicitation are part of our training.  Then our experts are by their side as they hone their skills through email, live-meetings, a 24/7 hot-line, and on-site visits.  Throughout we provide people with benchmark measurements of other similar efforts and tips to guide them. Suddenly fuzzy requirements become clearer. People are doing more accurate estimates in less time. Typically our clients improve their estimating accuracy by 50% within three months and several of our clients now produce estimates that average within 5% of actual scope, cost, and schedule.  Our clients leverage our cost estimating software and cost management software to estimate software costs and schedules, IT infrastructure costs, and custom hardware to improve their overall project and portfolio management.  By the way, I just finished a mile in the pool and improved by another 10% - practice makes perfect.

As I prepare my remarks for the first PRICE Systems International Symposium and User Group Meeting in Asia hosted by the PRIGENT corporation, I am astounded by the recent globalization of the Defense Industry.  Worldwide weapon systems acquisition has been permanently changed by: 

1. the merger of US defense contractors in the 1990's,
2. the entry of European contractors into the US Defense Industry (EADS, BAE) in the 2000's, and
3. the entry of Korea into the global market happening now.

Today, BAE Systems is the sixth largest U.S. defense contractor, U.S. defense contractors dominate European Defense top-ten lists and S. Korea is building its global customer portfolio, including Turkey. Among of the greatest challenges to this newly globalized industry are communication and understanding between customers and contractors, and among the internal organizations within a multi-national corporation. Defense procurement agencies and contractors wishing to succeed in this new world must adopt new communication strategies. 
One of those strategies will be evident at our event.  Negotiations to set the right price and to determine the ensuing operation and support costs are difficult in this environment.  Cost models, which work in the precise global language of mathematics, are the Rosetta Stone in the procurement Tower of Babel.  At our Symposium TruePlanning and the PRICE models will be the common language among Defense procurement professionals from around the world. And they will have little trouble communicating and understanding because the models distill their differences down to a universal set of cost drivers, activities, and resources common to all projects. We will discuss estimating software costs, IT project cost estimating, the effect of software development tools, the challenges of project planning and budgeting - and there will be understanding. When we used acronyms to name our model parameters, an arcane term like ECMPLX (engineering complexity) became the global term for describing the technology maturity and engineering experience of a development task.  Today, our clients from around the world have a universal understanding of our model parameters and can communicate and negotiate effectively using our cost estimating software.  Everyday, I see evidence of vast productivity improvements when multi-national companies and international procurements standardize on our models and cost management software, establishing a common language in the midst of diversity. 

Is it like a Tower of Babel where you work? Consolidation is occurring in nearly every industry today, so think about the efficiencies that can be gained through standardization?  ...communication?  ...immediate understanding?  As you improve communication and understanding, you will become more productive and estimate more accurately.  

While many at Bear Stearns, Lehman Brothers, Merrill Lynch, and AIG are staggering from the Wall Street financial crisis, others on the acquisition side are contemplating the fate of redundant operations created by the these consolidations.  This reminds me of the rapid consolidation of the global Aerospace and Defense (A&D) during the 90's after the fall of the Berlin Wall.  Granted, this consolidation is happening at a much faster pace, but it would be wise for decision-makers to do a little performance measurement before they slash and burn redundant operations. 

During the A&D consolidation, the General Electric Corporation called upon me to help them measure performance of several redundant operations.  Why me?  Well our activity-based estimating models do a tremendous job of normalizing and measuring the productivity of similar, but somewhat different operations.  When you provide costs as an input to our models, you can calculate productivity.  Doing this for several projects at different operations gives you a clear picture of which operation is the most productive. Decision-makers can then make unbiased decisions about which one to close.  The same method can be employed with our TruePlanning for IT estimating suite.  Centers of excellence and laggards for IT budget planning, IT project management, software development, IT data-center operations can be quickly identified and the best allowed to survive. So, let's hope that the decision-makers are doing their due diligence and apply proper corporate governance.  A little bit of performance measurement modeling will go a long way.