To follow up on my previous post (“Is Finkelstein totally clueless about the Internet”) with real data, I examine the stats of the log files on my server.

Below is a table generated by the log file analyzer Awstats for the first 2 months of my server http://landshape.org.

Month | Unique visitors | Number of visits | Pages | Hits | Bandwidth |
---|---|---|---|---|---|

Jan 2012 | 7,361 | 18,526 | 71,689 | 204,718 | 3.02 GB |

Feb 2012 | 7,081 | 16,422 | 113,111 | 233,158 | 7.67 GB |

You can see the number of hits for January and February is 205K and 233K respectively, and the number of visits is 19K and 16K, about 10% of the number of hits.

The number of unique visitors in each month, that is the number of unique IP addresses that views of the blog originate from is 7K, or about 40% of the number of visits. This would be the best indication of the number of possible readers of the blog.

But this still exaggerates the number of readers, as many people land on the pages from search engines, recognise its not what they were looking for, and click away almost instantly – the ‘blink’ effect.

Below is a table of duration of visits, where it can be seen that 79% of visits last for less than 30 seconds.

Duration of Visit | Number of visits | Percent |
---|---|---|

0s-30s | 1,322 | 79.4 % |

30s-2mn | 48 | 2.8 % |

2mn-5mn | 21 | 1.2 % |

5mn-15mn | 31 | 1.8 % |

15mn-30mn | 36 | 2.1 % |

30mn-1h | 46 | 2.7 % |

1h+ | 121 | 7.2 % |

Unknown | 39 | 2.3 % |

Therefore the total number of effective readers per month on http://landshape.org is approximately 0.1*0.4*0.2 = 0.008 or close to 1%. So my guesstimate from yesterday was pretty damn close. The ratios on other blogs may be a little different, but not so different as to matter.

Case closed your honor. Your proposal to regulate blogs with more than 15,000 hits per annum or 1250 hits per month would impact all blogs with more than 12.5 readers a month, or less than one per day.