Posted by George Huhn on Mon, May 17, 2010 @ 02:27 PM
What does it mean when a meteorologist says "the chance of rain today is 60%?"
Each day in the United States, a massive amount of data is collected from weather stations, satellites, and weather balloons from around the world and sent to the National Meteorological Center near Washington, D.C. The data is processed to give a multi-dimensional picture of global atmospheric conditions, and then it is analyzed using various algorithms to develop local weather forecasts and predictions.
But this isn't how they make the "percent chance of precipitation" predictions. Even with the massive amount of data and super computer speed, their predictive algorithms alone just aren't good enough. So they use comparisons to historical data.
Basically, they take the current atmospheric conditions and compare them with days in the past that had very similar conditions. So when they say that "the chance of rain today is 60%," it means that it rained on 60% of the days in the comparison set.
And guess what? Assuming the data was entered properly, these predictions are 100% reliable all the time. Why? Because they are only predictions of probability – they aren't "wrong" on a particular day, whether it rains or not. But whether they are accurate or not in the long term is an entirely different question.
The only way to determine if the predictions are accurate is to collect the data and plot the actual versus the predicted conditions over time to learn the margin of error. If it only rained on 30% of the days that the prediction was 60%, then there is a problem with the data or the data processing.
You can do the same type of probability prediction and testing with your business projects, too. The more accurate your estimates, the more confidence you will have in your overall project-value ranking in your project portfolios.
Developing more accurate project risk estimates requires 4 basic activities:
1) Identifying the key drivers of cost, time, and resource risks in completing project tasks.
2) Preparing a database of these tasks that includes the corresponding cost, time, and resource estimates assigned to each project and the basis for those estimates at the beginning of the project.
3) Tracking the actual costs, times, and resources used performing the task as each task is completed.
4) Comparing the actual costs, times, and resources with the starting estimates.
After you have maintained this database for a period of time, you will be able to plot the actual versus the predicted results. This plot will show you the accuracy of your cost, time, and resource estimates as well as revealing the distribution of the actual results. (You will probably learn that your cost estimates were too low, your time estimates were too short, and your resource estimates were for too few. And that is a good thing to learn.) Eventually, you will be able to use the actual results data as a basis for future probability predictions, including understanding the uncertainty in those estimates.
I saw the data of one major pharmaceutical company who did this for their project "percent probability of success" estimates. The data between 20 and 85% was surprisingly linear; for example, about 50% of the projects that had "percent probability of success estimates" of 50% were ultimately successful. It also showed that all projects that had an estimated "percent probability of success" of 85% or greater succeeded and all that had an estimate of 20% or less failed.
If you’re involved in project portfolio management and you're looking for ways to improve your project planning, compiling and analyzing your historical data is a great way to test and improve your future estimates.
Does your company track and analyze historical project management data? Why do you think that most businesses don't?
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Fri, Apr 30, 2010 @ 10:38 AM
This last paragraph caught my attention:
Karl Popper, the great philosopher of science, once divided the world into two categories: clocks and clouds. Clocks are neat, orderly systems that can be solved through reduction; clouds are an epistemic mess, “highly irregular, disorderly, and more or less unpredictable.” The mistake of modern science is to pretend that everything is a clock, which is why we get seduced again and again by the false promises of brain scanners and gene sequencers. We want to believe we will understand nature if we find the exact right tool to cut its joints. But that approach is doomed to failure. We live in a universe not of clocks but of clouds.1
Our businesses are mixtures of clocks and clouds, but it is often difficult to distinguish which parts are clouds and which parts are clocks. The clocks are things that we can measure and control but clouds are things that we can only try to predict.
Confusion begins when we try to measure and control clouds as if they are clocks or we focus only on trying to control clocks without trying to predict and mange the clouds.
Like Leher's comment on science, it is also a mistake of modern business to try to pretend everything is a clock. While I am a strong proponent of using business analytics, such as
simulation and
optimization in
project portfolio management tools, the analytical tools that we use should be able to manage data from both clocks and clouds. And when we're using these tools, we should be ever mindful of the inherent uncertainty (cloudiness!) in the data and the resulting predictions. Business analyses should rarely be
chiseled in stone. Too often, they are treated like they are.
Look around your business today. Where are your clocks? Where are your clouds?
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Fri, Apr 09, 2010 @ 09:32 AM
"It’s like we said on the iPad: if you see a stylus, they blew it. In multitasking, if you see a task manager, they blew it. Users shouldn’t have to ever, ever, EVER think about that stuff."
I love that quote. Why?
Because it shows why Apple is uncompromising in trying to bring their customers the best user experience that they possibly can. It is exactly the reason that they don't launch mediocre products, in spite of media critics who think, for example, that Apple should have launched a net book or are impatient for features that other phones have, but that are poorly executed. They won't do something the way everybody else does it when they believe that they can find a better way.
We try to take the same approach in developing business software applications. For the most part, our customers are not programmers; they're not operations experts; and they're not mathematicians. But they are really smart business managers who want to use business analytics to solve important high-value problems quickly. They want to think about getting to the solution – not how to write equations or how to integrate applications or how to program a spreadsheet.
For example, when we design a
form we have certain criteria that must be met, such as:
- Minimize redundant data entry by using form objects such as drop-down menus
- Enable and disable controls based on the users selection
- Avoid the need to display "Error" screens – if we come to a point where a user might get an "Error" screen, then we try to see how they got there and how we can prevent it.
- Make sure a full description of each form's functions is one click away.
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Thu, Apr 08, 2010 @ 10:42 AM
Not quite, although that was the promise in the title of an
article in the March 2010 "Wired" magazine about a recently invented algorithm called "Compressed Sensing" or CS. That isn't to say CS isn't a really cool algorithm for certain applications - it is and you should know about what it can do – but it can't make something out of nothing.
To understand the concept behind CS, think of a digital image. Different kinds of compression technologies (jpg, gif, png, etc.) are used to shrink the size of these images so they use less memory to store and process. Basically, these technologies work by using clever ways to reduce redundant or repetitive data points to a much smaller number of data points. For example, a large area of a single color can be saved without saving each data point since the color is identical. But even though compression can reduce an image size significantly, the compressed file still holds all of the essential information to display the image in its original detail and resolution.
Therefore, a digital image that is compressed to, say, 10% of its original size has essentially discarded 90% of its original data as unnecessary. So if 90% of the data that was originally collected by the image sensors is unnecessary, why collect it in the first place? Why not just collect the essential 10%?
The idea behind CS is that you can collect digital image data using far fewer physical sensors than would normally be used and then use the CS algorithm to reconstruct the digital image as if you had used a conventional number of sensors. So you lose the computationally expensive overhead of collecting all the data, analyzing it, and then discarding most of it. Instead, you only need to collect a small amount of it, and then use CS to reconstruct the rest. And CS can do a remarkable job of reconstructing an image from very little data.
The CS algorithm isn't just applicable to digital images. It can be used on all kinds of digital data processing from music to interstellar radio waves to scrambled radio communications.
CS works based on a concept called "sparsity," which describes the density of data. Conceptually, a floor that has a few balls spread out over it would be considered sparse whereas a floor covered with many balls of different colors all touching each other would not. It turns out that reconstructing an image using CS means finding the sparsest image that can be constructed from the dataset.
However, there is one key point that needs to be stressed: CS cannot reconstruct data that isn't there - you can't make something out of nothing. In other words, if you take a digital image using far fewer sensors than normal and there is a critical detail that is missed entirely by the sensors, it cannot be recovered using CS. Unlike CS, conventional image compression works well because it looks at all the detail first and throws away the data it doesn't need.
Nevertheless, the promise for CS is exciting, particularly in areas where full data collection can be difficult or impossible because of volume of data or physical constraints. These can be sampled using far fewer sensors than otherwise might be required and then reconstructed using CS to obtain a resolution that is adequate for extracting information. One of the major challenges in applying CS is determining the minimum number of sensors required to sample a given data set.
I have been thinking about how CS technology might be applied to business, process development, and manufacturing data. Can you think of any potential applications in your business?
Post script: The original title in Wired magazine was "F_ll _n T_e Bl__ks: A revolutionary algorithm can make something out of nothing." The on-line version's title was changed to "Fill in the Blanks: Using Math to Turn Lo-Res Datasets Into Hi-Res Samples." Much better.
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Mon, Jan 04, 2010 @ 02:08 PM
If you don't know the values and costs of not executing your projects then you're probably not maximizing the value of your project portfolio and you may be working on the wrong projects.
When project portfolio managers meet to decide which projects that their businesses are going to execute and which they are going to reject, they often have a summary business case for each project that includes the business value and attributes. Business attributes can include selection criteria such as
net present value (NPV),
return on investment (ROI), costs, resource requirements, and risks.
Thus, when the managers select a project to execute, the value and associated costs of the project are added to the total portfolio value and costs, respectively. When they reject a project, usually the identical "if-executed" values and costs are subtracted from the total portfolio because there is no separate evaluation of the value and costs of not executing the project. Therefore, the value of a rejected project is essentially set to zero by default and the total portfolio loses value.
When they reject a project in this way, any intrinsic positive or negative values and costs derived from not executing the project are not factored-in to the final portfolio. And when these values and costs are not factored-in, the total portfolio value and cost can be dramatically over- or under- estimated.
There are many ways a project can add or subtract value from a portfolio. Even projects that have negative individual ROIs can add value, such as a project that adds revenue to a product line because of its strategic fit. Analogously, there are many ways that not executing a project can add or subtract value from a portfolio. For example, positive value can come from increased revenue streams if the rejected project would have cannibalized revenues from other products; and negative value can come from a loss of revenue from a product line that could have been enhanced by the executing the project. Costs that can be incurred from not executing a project might include costs associated with contract terminations, closing facilities, and reassigning resources.
So, perhaps counter-intuitively, you can see that rejecting (not executing) a particular project may actually add more real value to a project portfolio than selecting another project!
How can you ensure that you're capturing the value and costs of not executing a project?
For each potential project in your portfolio, you could create an associated "Not" project that includes the overall value for not executing the project calculated using the identical attribute categories (rewards, costs, risk, etc.). Then, before
optimizing the portfolio against constraints, you could set up a mandatory dependency between these two projects such that either the actual project is selected
or its corresponding "Not" project is selected. In this way, either the value and costs of executing the project OR the value and costs of not executing the project are included in the portfolio totals.
Of course, if the value and costs of not executing a project are truly "0" and do not impact the total portfolio value and costs, then you don't need to create an associated "Not" project.
In our
project portfolio management tool Optsee®, you can perform rigorous
project portfolio optimizations against multiple constraints (such as limited money and resources) while maintaining four different types of project dependency relationships, including an "Or" relationship. When you select the "Or" dependency relationship between two projects, either one project or the other (but not both) are included in the optimized portfolio. This way it is easy to set up and accurately analyze the real value and costs of your portfolios under different constraint combinations because you're factoring-in the values and attributes of both selected and rejected projects.
Do you currently assign values and costs to not executing projects in your project portfolios? What other suggestions do you have for capturing these values?
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Thu, Nov 19, 2009 @ 12:35 PM
Quantitative financial analysts ("Quants") who design trading algorithms or optimize them to work just a few milliseconds or even microseconds faster are in high-demand these days. Successful trading algorithms often work by taking advantage of millisecond market inefficiencies that aren't widely recognized by other traders. This advantage is lost once other traders discover and begin to trade on the same inefficiencies.
Since the profitable lifespan of proprietary trading algorithms keeps getting shorter, optimizing an algorithm to execute faster is one way to extend its lifespan. So the competition for hiring really good quants that can discover, develop, and optimize new trading algorithms is fierce.
Your business probably doesn't require microsecond decision-making, but the speed and quality with which your business can execute is going continue to become increasingly important. World-wide competition is speeding up innovation, shortening product development times, and reducing product lifecycles.
While many companies are using IT to improve business processes, most of them are just scratching the surface when it comes to using meaningful business analytics to speed up those processes.
When everybody in an industry is using the same IT and business processes, nobody has an advantage. The advantage will come to those who can discover, develop, and optimize business processes beyond what everybody else is doing. That might mean just looking at the little things that everybody else overlooks (or doesn’t think is important) and speeding them up by a few days or even just a few hours. Just as saving a few microseconds is important to a quant, saving those few days or hours could mean the difference between leading in your business or just being one of the pack.
Look around. What's limiting the speed of your business?
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.
Posted by George Huhn on Thu, Nov 12, 2009 @ 02:27 PM
Researchers at the University of Münster in Germany have shown that people can be highly influenced in selecting financial assets simply by the way the risks are charted. Their study, published in the June 2009 issue of Decision Analysis, showed that people are significantly more likely to select data charted in one way versus the same data charted in a different way. They also showed that people who self-describe themselves as "risk averse" (as most people do) will prefer one type of data presentation over another type.
In particular, two types of commonly used data distribution curves were discussed: probability density functions (skewed bell curves) and cumulative distribution functions (S-type or logistic curves):
In multiple experiments involving subjects selecting financial asset models using data with "right skewness" in probability density plots, people were significantly more likely to select the model if it was instead displayed as a cumulative distribution function (S-type curve). For financial asset models data with "left skewness," the opposite was true; people were more likely to select it if the model was presented as a probability density plot (left-skewed bell curve). Furthermore, the researchers found that "Individuals that judge themselves as more risk averse show a stronger preference for right skewness."
The authors note that:
"The findings of this paper have several practical implications. First, in advertising their products a financial services firm may choose the presentation format that is most likely to induce specific preferences for the considered financial product. For example, the sales brochure of a mutual fund investing exclusively in growth firms may use a presentation format that induces a preference for right skewness. Similarly, a firm advertising a discount certificate, a financial instrument that implements a covered call strategy and thereby generates a left-skewed distribution, may use a density function to communicate the asset’s risk."
What are the implications of this if you're not an advertiser trying to pitch financial services to potential customers?
Well, if you do any kind of business analysis presentations, like project portfolio management analyses, you'll want to remember that the chart types that you select for your data can strongly influence your audience based on their unconscious preferences.
And if you're looking at charts in order to make a decision based on the data, be sure that you look at the data in different chart types so you don't make your selection based on your own unconscious preferences.
What are the best uses of your company's dollars and resources? Optsee
® can tell you. Optsee
® is a project portfolio management and budgeting optimization tool unlike any that you've ever seen.
Click here to find out more.