the general trend is probably correct - social websites are gaining moment. still i don't like these graphs. we can not measure visitors (only devices, but you

the general trend is probably correct - social websites are gaining moment. still i don't like these graphs. we can not measure visitors (only devices, but you can do some statistical approximation - that's what they do for sure). but this gets even vaguer if you are not the owner of the site and don't have total access to all data. we don't know how google calcs those "numbers", not even how they get approximated (they are not really statistics - Y axis is not labled). it could very well be - and probably is, i've been in the webstats buiz - that the way they come up with this is by adding some artificial, best-guess multiplier to correct for not having enough real data. --- the way to do those web statistics right is: * have an open consortium with all the major players come up with an algorithm that satisfies them [1] * apply that algorithm, it will output numbers that are in some correlance to the real, but unobservable truth [2] * more important: the statistics of those websites are now comparable to one another. because the algorithm & factors are the same for everyone [3] * rinse, repeat, adapt [1] it is unlikely that they will publish this algorithm, since it will contain strange, seemingly arbitraty factors like "multiply by 1.2 during summertime, because people are less online and we don't want tooo much of a down curve" [2] like "visitors" or "time spent on site" [3] no, we don't know that's the case with google's data