A somewhat tongue-in-cheek guide to see how much time you want to spend investigating this product and how you might categorize yourself as a client.
Hello, Partner!If you fall into this category, we could ask for and get a piece of your company for licensing fees. Send your license fees today before we learn exactly how desperate you are. We're nice people but it is better not to tempt us.
If you are running Gosset designs, you probably don't have any other option. Currently, only STRATEGY for Windows can deal with the constraints that many Gosset designs require. (See Dave Doehlert's article: Why buy an Optimal Experiment Design?)
You are probably also in this category if you want to analyze large numbers of factors and responses. We absolutely guarantee that you will run into your budget limit long before you have an experiment that is too large for STRATEGY.
If you have a bet the business experiment on the line, can you really afford not to look at this product?
Career MoveThere's a lot of talent coming in this field. You need to stay ahead of the new talent, or you are the new talent and you need to make a reputation.
You see that being known as the person who solves the tough problems is your route to money and power, or maybe you just like the tough questions You've already successfully designed and run several experiments. They've worked out well, but you know you can do more. That means that you need software that handles more factors, more responses, and more complex constraints. In short, you need software capable of finding better answers.
Are you going to get passed by? Call today for your free evaluation copy to keep yourself ahead of the field.
Throw Me a RopeYou've already done your experiment, and it was very expensive. Unfortunately, you did not find the answer you were looking for even though all of your experience and intuition tell you that the answer is buried in the data somewhere.
Every day, experiments are abandoned as failures when the real failure is in the analysis of the experiment. The right answer is often there, but it is invisible without the right tools. Our competitors offer two types of tools for finding the ideal factor settings: contour plots and mountain-climbing algorithms. We can do better. (Check out our free experiment analysis offer.)
Contour plots are simple and effective for two-factor experiments. Add a third factor and finding the sweet spot becomes tricky. Add a fourth, and you quickly get to feel like you are looking for a needle in a haystack.
Mountain-climbing algorithms were utilized to overcome the weaknesses of visual inspection. This approach was very popular in the eighties because it runs very fast even on slow computers. Unfortunately, mountain-climbing does not always find the best answer even when combined with sophisticated random seeding strategies. The more complex the landscape (more factors), the more likely that the best answers will be missed.
We recommend and use GridSearch. It's slow, but it doesn't miss ANY good answers. The response values are calculated for EVERY combination of factor settings, and it works with any number of factors and responses. Besides, slow means something different these days. A typical GridSearch that took 100 hours on the original IBM PC runs in a few seconds on a Pentium Pro.
Buy our Competitor's ProductYou think Taguchi is the state of the art in experiment design. If you've gone beyond Taguchi and started looking at interactions, the idea of three, four, or more factors in an experiment is still a little scary.
If you are running simple experiments that don't cost very much, you can get by with just about any of our competitors. If you are running main effects designs or just looking at two factors, you can do a pretty reasonable job with any of the hundreds of tools that will run a standard regression analysis and generate a contour plot. It will be hard to justify the cost of our product, particularly if your organization has already provided a general purpose statistical tool with a DOE add-on component.
When your experiments become complex and expensive, think about STRATEGY for Windows. Until then, keep your wallet in your pocket.