Let's say you are thinking about running, or you are planning, an Archaeological Experiment. You have started with defining and phrasing your key question*, and you have done all the background research. Your outline is written and has been checked by someone else to help make sure that nothing is missing. You have made a list of the tools, materials, and other resources that you will need.
And now the actual experiment is drawing nearer and you have to prepare it. And do it. So here are the Rules for Experimental Archaeology...
Rule #1: Things always take longer than you expect.
This does not only apply to researching and writing the outline, but also to sourcing your tools and materials. And, of course, to the experiment run itself. Oh, did I forget the post-experiment analysis, synopsis, additional research and writing of the publication? That, too.
Rule #2: Something that you need is always missing.
If nothing is missing from the list you have written down beforehand, you will probably need something that was not on that list. Or you would have needed it twice and it's only available once. Or something is not working and therefore missing in regard to all practical purposes.
Rule #3: The only way to make sure things will not go wrong badly is by having enough surplus material available.
If you have just enough to make the experiment run... you are tempting fate. If something goes wrong and you lose some material, you will be right back at Rule #2, and in the worst case that will kill your whole experiment run.
Rule #4: You cannot document too much.
While having a huge load of photos and taking copious notes of everything (including subjective stuff) can be a real pain in the neck when finishing up the documentation and writing the paper, it's always easier to pick a sample of photos than do another run of the experiment because you lack documentation. Plus, in case of a repeat performance, one of those additional snapshots might be just the thing to remind you on how you did it last time... for example in case something is missing (remember?), and you have to find a substitute.
Rule #5: If you did not write it into your plan, chances are high you'll forget it.
Experiments tend to eat you alive, and they will suck up your brain capacity. It's even worse if you start the experiment run not really well rested (possibly due to Rule #1), or if it's a very long experiment (or becomes one because things take longer than planned), or a very complicated one. Thus it's immensely helpful to have a cheat-sheet, also known as The Plan, where you have written down all the single steps to do during the actual run, in order and with remarks of what you must document by taking measurements or photographs. A detailed plan is also a good way to make sure that everything is listed in your requirements list and a good way to estimate how long the single steps will take.
So... plan ahead, plan for more time and more material than you think you need, get someone (preferably with crafts knowledge in that area) or several someones to check your plan... and then go for it. Experimental Archaeology is a lovely thing to work in.
* If it has no key question that it's designed around, if it's not as objective as possible, or if it is not documented well enough to be repeatable, it's no archaeological experiment. It may be something else and it may be important or helpful... but no true experimental archaeology.