Quality Score is an incredibly important factor in managing Adwords campaigns, but is sometimes difficult to handle.

To understand how Quality Score is impacting an account it can be better to visualise it than stare at rows of keyword data. I’ve created a couple of treemaps to demonstrate how this can be done.

Example 1: keywords evenly weighted

Click the image for a bigger version

How to interpret the treemap:

• Each square represents a keyword and match type combination
• Keywords are grouped into adgroups, which are grouped into campaigns
• The colour spectrum of the box represents the quality score: 1 is deep red, 5 is white, 10 is deep green

So you can see that there are some problems with this account focussed in particular adgroups. However this view doesn’t give much of a sense of the impact of these quality score issues.

Example 2: keywords weighted by impressions

Click the image for a bigger version

The interpretation is the same as above, except now the size of each keyword box is determined by the number of impressions it receives. Suddenly problem areas become more apparent, particularly with the adgroups in the top left of the image.

This is a more impactful way of demonstrating Quality Score issues than dry numbers and percentages. For other ways at looking at SEM data with Treemaps, take a look at this post on the Efficient Frontier blog.

Figuring out new and interesting ways of representing paid search data is one my particular interests. I found a bit of time today to play with Google Charts API which has a great Venn diagram generator.

Although you can easily create Venn diagrams ‘freehand’ in Excel, producing them accurately from data is much more difficult. However, it didn’t take me long to whip up this chart:

The data shows which conversions come from keywords that are on just broad, phrase or exact match, or a combination of them. The end result is probably not too meaningful, but it was very quick and easy to make.

I’ve been researching other tools lately such as Many Eyes and Processing. Hopefully I’ll be able to post a few more examples soon.