Hi, Just curious how much traffic a website should have before it's safe to experiment with ads placement and colours? Obviously, the more traffic the better, but is there a number of daily visits or impressions which I should use as a guideline? Something tells me it's not worth much to experiment with a website which has no real audience - most results will be random and therefore make conclusions invalid. Thanks in advance for all the help!
I think it wouldnt hurt to put a few ads on but if its not got loads of visitors at the moment i wouldnt go crazy with ads also it might not work experimenting coz you wouldnt have anything to compare it to Becky
That's my thoughts exactly! Thanks, Becky! I do have ads in place, but try not to experiment too much with colours, unless it's needed to blend them into the new website theme. I figure 1000+ daily visitors and/or RSS readers is a reasonable enough start for valid experimentation. What do you think?
It depends where your visitors comes from. If you have 1000 visitors a day, all from digg and stumbleupon, you cannot rely on that becaue that visitors dont click your ads. If you have 100 a day from various sources, then you can experiment for sure.
It's difficult to judge the results of experimentation when you only have a small amount of traffic. You don't know if changes are the result of your experiment or to normal variations.
if i had 100 visitors a day to a site i'd start to experiment a little....but i wouldn't spend too much time it...just work on the content and marketing and drive up your traffic. Are you making some kind of revenue now?
That's it, content is my priority until I get a few thousand daily visitors. Otherwise even a small change may look like some big shift just because the overall number of visitors is not large enough.
I think at least 500+ visitors per day to test substantial changes - these should show up within a few days (although perhaps not for low CTR traffic like forums or Digg, as mentioned above) To test subtle fine-tuning changes that might change the CTR only a small amount (from, say, 3% to 3.3%) you need more visitors though. The more subtle the change the more visitors you need to make testing valid.
Rusputin makes a good point. With 100 visitors CTR can vary naturally from 2% to 10% one day to the next. So you can't be too sure if results are from natural variation of from your experiments. However, if you run the test over a long period the results might be more accurate.
I would say that 500. + should be a minimum... As others have commented on before that 100 is to unstable.. Whit a higher amount of uv you can change more as well many try out 3 different colors at the same time.. Whit whit hundred uv I do not hint that the final result will be "valid"