Test little things? Or test for Impact?

Some people say, go do a split test or multi-variatetest on little things. Get the results quickly, and adjust quickly.

Some people recommend, don't just test and test meaningless things, test stuff with impact to the whole site.

I have tried both.  I happily test little things like a headline here, a copy there, layout, pictures, etc.  It takes me a long time to think of a test that will impact the whole site, but when I do think of something, I am very excited.

So what do I recommend?  While of course testing a hypothesis that will benefit the entire site should be the reflexive answer, testing little things is also imperative?

It's in testing the little things that you gather the ideas to have a grand hypothesis that will impact the site or the sales process.  It's in quick and "meaningless" tests that you realize there is this low hanging fruit of a hypothesis waiting for you to test to improve your site.

So I suggest do both.  Constantly work to do testing with big impact, but do your little testings along the way and learn from it.  Constantly try to integrate the big and little learnings, and truly your website will benefit in the long run.

PS. Warning to the faint-hearted:
Just because it took you days to think up a hypothesis for the entire site, doesn't mean that your traffic will gobble it up and give you quick results.  You may be surprised no one will pay attention to it :-)  I'm just saying.


Want free updates from OFW Din?
Subscribe to Online Filipino Worker - OFW Din!