Completely agree. Sampling is fine; but you need to understand that the free version of Google Analytics isn't a replacement for a proper BI tool. GA also doesn't even do sampling until your traffic is over a certain threshold, so at low traffic levels it's actually accurate (though those levels are also low enough that you probably shouldn't try to draw too many insights).
This is not to say that GA is bad; in fact I would call you stupid if your web startup went out and bought a tool before you launched. GA is free, it works well, it's super-easy to implement, and people should use it; but once you've reached the point where you can afford something better (usually in addition to GA, not in replacement of) you should look into doing so.
I think Google Analytics is much better than nothing, and until you reach a certain scale (I used the estimate of ~1 million monthly sessions) I think it's sufficient.
What I am mainly arguing is that relying completely on GA for reports and performance measurements is dangerous and frustrating.
Comments
Completely agree. Sampling is fine; but you need to understand that the free version of Google Analytics isn't a replacement for a proper BI tool. GA also doesn't even do sampling until your traffic is over a certain threshold, so at low traffic levels it's actually accurate (though those levels are also low enough that you probably shouldn't try to draw too many insights).
This is not to say that GA is bad; in fact I would call you stupid if your web startup went out and bought a tool before you launched. GA is free, it works well, it's super-easy to implement, and people should use it; but once you've reached the point where you can afford something better (usually in addition to GA, not in replacement of) you should look into doing so.
I think Google Analytics is much better than nothing, and until you reach a certain scale (I used the estimate of ~1 million monthly sessions) I think it's sufficient.
What I am mainly arguing is that relying completely on GA for reports and performance measurements is dangerous and frustrating.
- Simon