SEO Speedwagon

In my mind’s eye I have an idealistic view of each paid search account that I manage. Of course it includes my clients achieving stratospheric ROI, but at heart I am a numbers guy. I know that in order to get the best possible results it requires the best possible data. However I do live with both of my feet on the ground and I realize that not all clients possess the ability to successfully deploy/integrate analytical tracking into their ‘web systems’ for a plethora of reasons.

The idealist in me wishes this were not so, while the realist tells me to move on. In that vein, and using the data that is readily available to all PPC campaigns there is still a lot of analysis that can be done to help guide us in making decisions. The hardest part is deciding which piece of the data that we do have should be ‘worth’ more than the others. In light of that, I try to visually represent things as much as possible so that they can stand or fall on their own merits and go from there.

To borrow a phrase from someone much wiser than me, I have found that where there is a will there is a way. Part of that way for me is described below in a section of a case study for a client that is not able to take advantage of analytics. Please keep in mind that this is not a complete analysis but rather highlighting how to find a way to gather useful data out the information that we do have.

Some Basic Facts:
•Who: A large client that has excellent brand awareness and a sizable spend
•What: Specific areas of interest are in brand vs. non-brand and key phrase length
•Where: Continental U.S.
•When: 3rd Quarter 2007
•How: Analyze the ‘big five’ data points (Impressions, Clicks, Click-Thru-Rate, Average Cost-Per-Click, Cost).
•Statistical Constraints: To be included in our data set each phrase must have had at least 1,000 Impressions and a Click-Thru-Rate of at least 5%.

Branded Terms

When I look at branded phrases the first thing that sticks out to me is the obvious dominance of the 4 word key phrase. In this instance the length of the brand name has a lot to do with this, also but notice that 1, 2, and 6 have roughly the same footprint. How do we make a judgment between 1, 2, & 6? Well, taking a more holistic view will aid us in this.

From the pie chart we can see that the 6 word key phrases rate far better than 1 or 2 word key phrase’s because of the negative impact of the higher cost associated with them.

Non-Branded Terms

When I look at the non-branded phrases I am first struck by fact that only key phrases with 2 and 3 terms qualified for inclusion in our data sample. The biggest difference, however slight, is in the impressions where 2 wins over 3.

The next fact that sticks out is how close the two phrases are together in overall performance. They may have individual nuances that can be tweaked but neither phrase should be overlooked in its importance to the success of the non-branded phrases.

Overall
The final piece that I will examine is comparing the branded phrases directly to the non-branded phrases. I find it important to introduce another measurement to the ‘big five’, number of keyword phrases, at this point to add weight to the question of reach.

The most striking thing besides the dominance of the branded keywords, is in looking at the enormity of the Average Cost-Per-Click for non-branded terms.

Conclusion
As a campaign manager one of my core duties is to make sure that I have collected as much important information as possible before we start to make decisions that guide the direction of the campaign. Although I would prefer a campaign with fully integrated analytics that is not always possible. In that case it is vitally important to make the best analysis that we can with the data that we do have.