When I first started using AdSense, I knew nothing about contextual advertising. It showed in my revenues. For the first few months I was making little more than a few dollars a day.

I had to go through a long learning curve, experimenting with formats, appearances and location. I had to test keywords and colors and fonts. I played around and gradually my revenues rose until eventually they reached the happy levels theyre at today.

When I think about how much money I could have made during that learning period, I get goose bumps. Im very glad Im not in that position any more.

And Im not too pleased about finding myself in danger of slipping back there every few weeks.

It seems as though the online advertising world is changing all the time. Every month, Google makes a change that could have an effect on revenues. Or some company brings out a new contextual advertising system. Or we discover some former secret about how a system were already using works and rush to figure out what we can do with that knowledge.

All of that requires more experimentation.

For example, Ive recently been playing with eMiniMalls. Ive come up with some pretty good ideas to make them generate maximum revenues but again, Ive had to go through a fairly long learning curve to figure them out. And while I was figuring them out, I had to make changes to websites that were already bringing in money with only a vague idea of what the effects might be.

I sometimes wonder if I shouldnt just put up a website that I could use specifically for experimentation. Instead of trying new strategies on sites that I knew worked, I could test them on a practice site first and then incorporate whatever worked into my revenue-generating sites.

That would certainly save me from effectively gambling with my current revenues as I try to figure out ways to use all the new tools coming online.

But of course Id only be losing all the revenues that that test site would be earning if Id optimized it properly.

Still, even if revenues do fall while Im playing with the location of an eMiniMall ad or adjusting the format of my AdSense units or trying to get different AdLinks to light up, the final result is usually a higher income later. I guess those revenue dips while we figure stuff out are just part of the capital investment we all have to make.