# Technorati Rank: In Depth & Explained 5 September, 2006 — Stuart Brown

## Meaningless number or viable benchmark?

Posted in Web Design, Analysis

Tagged with: technorati, rank, ranking, blog, blogosphere, blogs

Technorati has established itself as the definitive tracker for the blogosphere - with more than 50 million blogs in their index, they certainly have a gargantuan task at hand when it comes to organizing that information. Much in the same way that Google ranks search engine results by link popularity, Technorati keeps track of who's linked to who in the world of the blogosphere, accumulating data on the 'authority' of such publications. One interesting side effect of this is that Technorati can effectively 'rank' the entire blogosphere, based on link popularity - but what can this figure tell us?

Technorati rank checking can be addictive - watching your rank slowly climb up as you garner more links is a most satisfying process (almost as good as checking one's Adsense stats). As I was check Modern Life's rank for the *n* th time, I was inspired enough to look deeper into the rankings, and see exactly what the relationship between rankings and links really was.

Using the 'Top 100' blog listing, which lists the top 100 ranked blogs, along with their respective links, I was able to gather a snapshot of the most significant blog to link ratio data. Here's what a plot of that data looks like:

### Analysis

So what does this graph mean? Well, it indicates the obvious - you need to get more links to get a higher rank - but look at the links the top 10 blogs - you can see a drastic rise in the number of links required to break in here. Note also the 'long tail' - a term popularized in recent years, but particularly applicable to the blogosphere.

Only the top 100 blogs are shown, but a rapid dropoff in the top 10 is followed by an increasingly slow descent. Mathematically, the curve will never reach zero - but in the real world, the discrete nature of inbound links means that the a plot of the figures would only hit zero at around the rank of **1.5 million**. Imagine extending the graph above by a factor of **fifteen thousand** across - now *that's* a long tail!

### Formula

The curve of the graph can be defined with mathematics - the graph can be modeled quite closely with a power law function. The power in question very closely approximates *-0.5* (-0.51 was the figure I came out with), which is the inverse of the square root.

We can express this as follows:

Where l is the number of inbound blog links, r is the technorati ranking, and c is a constant.

Calculating c is rather simple - if r = 1, then c = l. In other words, c is equal to the number of blogs that link to the top ranked blog. In this case, it's around the 26,000 mark, but we can approximate this to 30,000 (in actual fact this models the graph better at the lower rankings). Hence:

We can also rearrange the equation to give the more useful equation to calculate ranking, given the number of blogs linking to the site in question:

So, as an example, Modern Life currently has **373** blogs linking to it. Using the approximation for c, the function gives an estimated rank of **6,468**. The actual rank is around the 6,000 mark, so the function does give a reasonably good estimate of rankings. Minor caveat though - the figures drift out the further from the top 100 you go, so lower rankings are likely to be far more inaccurate.

Nevertheless, it does give an interesting insight into the maths behind the blogosphere - and a good estimate of how many links you'll need to get in order to become a top 100 blogger.

Ranking | No. Of Links |
---|---|

10 | 9,238 |

25 | 5,789 |

50 | 4,065 |

100 | 2,854 |

250 | 1,789 |

500 | 1,256 |

1,000 | 882 |

### Rankorati

Just for fun, here's a small Javascript applet that uses the above functions to provide an estimate of the ranking / links relationship. Type in the rank or number of links, and click the respective button to calculate the corresponding figure!

Should you wish to, you can embed this applet into your site with the following code snippet: