If like me, you work better with visual representations of data rather than looking at columns of numbers, Google’s relatively new Fusion tool can help. It shows you the relationship between all sorts of data in a visual way, sometimes revealing previously hidden relationships.
In the following graph, which takes just seconds to make, you can get a real feel for the quality of the link graph around you main anchor text. For example, “Italian recipes” has a great, diverse incoming link graph but “great Italian recipes” less so.
With the length of the line representing the citation flow or value of the link, you can also see some useful outliers.
How can you create graphs like this? Very easily with Google Documents.
First, you need some data. Go to your favourite link data tool, (Raven Tools, Ahrefs, SEOmoz) and grab some data for your site. For a simple start, get a list of incoming links (you can use up to 10,000) and and their associated weight or PageRank, MozRank, or other metric.
Put these into a Spreadsheet thus…I’ve quickly added a first column identifying the domain that the Source URL links to. It would be even more useful to get the actual URL that the Source URL links to.
Next, go back to Google Docs / Drive and select New Table (beta.)
When asked to choose to import new table, select the spreadsheet you just created (you can upload a csv, txt or tsv file or select a Google spreadsheet.)
You’ll be asked where the column names are and what you want to call the table, and you are nearly there!
On the next screen select Experiment > Network graph. Immediately you’ll start to see a data visualisation, that might or might not be useful.
This is showing the incoming links to the target site, with the link length representing the value of the link.
By adjusting the network visualisation settings top left, you can get more interesting data such as a view on incoming link strength and source by anchor text.
With more detailed data, such as the source URL and exact target URL, you can build some really interesting and informative network graphs.