Experimentation

Experimentation can enable you to accomplish some great things. When I talk about the concept of “experimenting,” I’m referring to the process by which you test and then refine ideas and processes.

For example, my fiancé and I were recently working on wedding invitations. We decided to print all of the invitations at home rather than pay for an expensive service, but we weren’t sure exactly how to do this. So rather than try to get everything exactly correct the first time, we just did a test. We held the expectation that something would likely be incorrect on this first attempt, and sure enough it did not print with the content correctly centered. But we took the results of that test and refined our method several ways and tried again. After one more attempt, we were completely successful. This was a very quick and easy process, and nowhere near as difficult as we could have made it for ourselves if we had tried to get everything perfect the first time.

I could have also called the title of this article “Iteration.” Because that is exactly what I’m talking about: when you repeatedly test and refine your process or idea, this is equivalent to “iterating” on that process or idea. Wikipedia defines iteration as “the act of repeating a process usually with the aim of approaching a desired goal or target or result. Each repetition of the process is also called an ‘iteration,’ and the results of one iteration are used as the starting point for the next iteration.”

This concept can be applied to almost every aspect of your life. When you write a paper, typically it is best to have several people edit it in order to enhance and clarify the content. When you code up a complex software simulation, you should test small chunks of the code at a time. By testing each block of code before you attempt to test the entire program, your chances of success dramatically increase. When you’re cooking a meal, typically it’s best to test as you go along, to ensure that the final meal will taste best.

When you’re doing something for the first time, like learning how to use a power drill, often the best way to figure out how to use it is to try the different settings (in a safe way obviously) until you get what you want. I often call these scenarios “running an experiment.” This makes it feel more like a fun activity than a frustrating set of failures. Failure typically has a negative connotation, but it is the best teacher you can have. There is a great deal of content available online and in many books about the concept that “failure is your friend.” Learn to embrace it and you will improve at a rate that you thought previously impossible.

Another approach that many others have written about is “Ready, Fire, Aim.” This is the idea that instead of attempting to hit your target perfectly on your first try, it is better to shoot off that arrow first (especially if those arrows are cheap) to see how the winds are behaving, and then use that information to fire much more accurately.

Another way to phrase this concept in engineering language: you should continue to iterate in a while loop until you reach your convergence criteria. The cost function that forms the convergence criteria may be challenging to define, but typically in engineering problems there is a “good enough” threshold that will allow the iterated product to perform it’s job sufficiently. For those who don’t speak engineering: you want to keep improving your process, idea, product, etc. until it’s good enough for the purpose you intend. I can hear the comments now: “Well why didn’t you just say that in the first place?” The answer is that engineers like to show our technical superiority to other human beings by coming up with lots of fancy words that others don’t understand. :-) Either that or we’ve been staring at code too long.

Here’s another example from recent experience: I gave a dry-run of a presentation for a conference next week at my office, which lasted about 20 minutes. The audience members then spent about 10 seconds giving a few complements regarding the delivery and in the following 40 minutes they ripped apart almost every aspect of the presentation. Even though most really liked it, they had comments and suggestions for almost every slide, and as a result my slides are significantly better. Without this iteration, my presentation would be significantly weaker right now.

There are countless other examples. Building a treehouse, deciding on what products you want to sell in your new business, choosing a chair for your office, learning an instrument, etc. So the next time you are agonizing about a decision or how to get started on a large project, ask yourself this: can I run an experiment? How much will that cost, comparatively? If the answers are yes, and relatively little, it’s time to get experimenting!