Analyzing Data to Lower Your CPA

The optimization of pay-per-click (PPC) campaigns is often a challenging task. There are many considerations, and having great tools on hand, rolling up your sleeves, analyzing the data, and making smart decisions are all key to success. Today’s case study examines a few aspects for consideration.

Too many companies don’t have the time, budget, or knowledge, to properly implement their PPC campaigns. So they put up campaigns simply to test things out and see how it works for them, and the results are less than optimal. Unfortunately, the probability of this working is really quite small. Even in a non-competitive market, there are many ways you can get clicks on your ads that you don’t really want.

In a competitive market, the situation is far worse. Once there is competition for keywords, all chances of finding easy money with your PPC campaign quickly evaporate. If you are going to succeed, you are going to have to do some hard work. To illustrate, I will outline a case study provided to me by Brian Lewis of Engine Ready Software.

Digging Into a PPC Campaign

Engine Ready took on PENSRUS Corporation as a new client. When they dug into the PPC campaign PENSRUS had been running, the average cost-per-acquisition (CPA) was $150. Frankly, the cost was too high, and PENSRUS CEO Henry Morgan was about to give up on the campaign altogether. However, PENSRUS hired Engine Ready to give the campaign one more try.

The next step was to put into place Engine Ready’s Conversion AnalystTM software to provide detailed analytics of the PPC campaign. In addition, the account was examined in detail, and one of the first things that became apparent was that a large percentage of the keywords in the campaign were using broad match. While this option provides the most exposure to web searchers, it tends to be the most expensive because it is not well targeted.

Broad Match Caveat

In principle, broad match is supposed to offer you a wider range of relevant keywords. For example, if you enter the keyword “widgets,” it will also match up your keyword with “blue widgets,” “green widgets,” and so forth. In principle, this sounds great. However, if you are bidding on something like “surfing” with broad match, and you are selling surf boards, you are also going to get matches with terms like “web surfing.” This clearly does not meet your intent.

To make matters worse, Google makes use of “expanded broad match” functionality, as reported by Michael Pieper of Semtek in this blog post about Unexpected Match Types. According to this post, he had a client who was bidding on “Hong Kong night life,” and Google’s expanded broad match caused this ad to be shown on searches such as “Hong Kong sex,” which was not what the client wanted.

The bottom line is that broad match is a great way to get wide coverage for your PPC campaign, but it can bring in lots of traffic that has little chance of converting. One of the ways to address this is by using broad match less often, or being very diligent in tracking what search phrases are bringing traffic that you don’t want to your ads, and then using negative match to disable the keywords that don’t apply.

This is exactly what Engine Ready did for PENSRUS, and the results were very positive. That $150 cost-per-action was reduced to $80. Suddenly, the client was very happy with its PPC campaign, and was looking for ways to invest more — a big shift from being on the verge of giving it up.

For more information on the PENSRUS campaign, you can check the Engine Ready case study on their site.

Related reading

Simple Share Buttons