Google+ is NOT failing!
Last week I analyzed the data that I collected from Alexa and wrote a post Google+ is failing. Many people have pointed out that Alexa data is skewed and not reliable. This is true when comparing two different sites. However I believe that comparing Alexa data of one site over period of time will give some indication into the user volume trends of that site. Specially in the case of large sites like Google. As one commenter put it:
You can make the Alexa is only for people with the toolbar argument when comparing two very differnet sites. But not when you are comparing the SAME site over just a month apart unless you can come up with a reason why 30% of Google+ users who had the toolbar installed (and were therefore being counted) all decided to delete the toolbar now. – Lakawak
Another possibility would be that Google.com overall traffic has surged enough to decrease Google+ relative percentage.
…there is NO WAY you could think that Google’s overall traffic would increase to the point of making up for the 30% decreasing… – Lakawak
Now these points out of the way, there is at least one reasonable explanation that accounts for declining number of visitors but does not mean that Google+ is failing.
To put it very simply, there are two kinds of Google+ users. True Fans are the users who love Google+ and will use it more and more. Hype Users are the ones who just signed up to see what’s this all about and have now moved on to next new shiny object. It is entirely plausible to assume that the drop in visitor numbers comes from hype users leaving the service while True Fans stay and their numbers actually increase.
[This is total fiction] To make it easier to understand I drew a simple sketch that illustrates the True Fan/Hype User point. The graph below is based on the current Alexa data and assumes (for illustrative purposes) that all hype users are gone by tomorrow.
I would really love to see somebody competing with Facebook, so I plotted slight exponential growth in there. However there is now way to tell if this is true or not. Overall numbers are going down and we have no data on different segments.
I hope this gives Alexa numbers some context.