SEM

You have just built a great Google Adwords, Yahoo! Search marketing or Microsoft adCenter campaign. Viewed from the web or local interface, the data is skewed towards analysis by keywords entered into the campaign. Assuming they are not set to exact match, this will not show the whole picture. It is important to dig a little bit deeper, and examine the phrases that trigger your ads.

When you look at the phrases that actually trigger your ads, such as in an Adwords Search Query Performance report, you will start to see patterns emerge, the most interesting being terms that are common across phrases that triggered conversions. Taking a closer look at Clusters of Queries like these can reveal a lot. Most SEM campaigns will have a subset of keywords that might not even be set up in the account itself that correlate strongly with sales. There will also be phrases or words that obviously do not fit or inflate the campaign’s CPC by crossing over into another industry with an aggressive SEM space.

Keyword performance will always be influenced by a number of factors. The effect will vary depending on other descriptive and qualifying terms. Some queries will naturally return more on investment, yield more traffic or be genuinely cheaper. Not all of this will be the product of consumer behaviour. Competition within the query space is incredibly important, especially once you start to examine the performance of specific clusters of search terms.

For example, suppose a pet shop called ‘I Like Turtles’ has set up a campaign to sell turtles. The landing page, with a strong call to action and a robust cart has been built, and a campaign built around phrase match on species terms like ‘Box Turtles’ has been launched. The traffic is not performing as well as expected. The CPC is unsustainable and the traffic is not engaging on the landing page at all. A quick look at the searches triggering the ads suggest a long list of additional negatives and reveals that some descriptive terms are performing better than others, while some are not even appearing at all. The same is true for locations. Reported first page bid levels can provide a guide to what is happening in the market, but it won’t reveal the reason.

What ads are triggered for a long search phrase?

There are a number of different ways one campaign can compete with others in the same query space by either bidding directly on the same keywords or bidding on other keywords that appear in ad triggering search phrases. There are a few that can create a skewed impression of the query space for any single account. Campaigns targeting specific areas within a larger query space for another coupled with specific geographical regions, devices and time periods can aggressively reduce the larger campaign’s visibility in otherwise profitable spaces.

Either through less wasted coverage leading to a better return or a more aggressive bidding strategy one can lock a more general campaign out of lucrative query groups. These competitors do not even have to be selling the same product to the same people; there can be a lot of overlap in terms used for the SEM campaigns of different businesses.

For ‘I Like Turtles’, their targeted query space may also have a toy store using “Buy Turtles”, travel and accomodation companies targeting the name of their city, a conservation organisation raising awareness on river conservation and a DVD store selling Ninja Turtles movies. These companies do not need to be selling the same product to have an impact on the ‘I Like Turtles’ SEM campaign. Travel companies may bid higher than the pet shop and be less precise with their negatives, and increase the CPC. The DVD store might time their ads to run for the weekend, and periodically bury the pet shop’s ads, and the conservation company and toy store may increase competition on terms that often form a part of a search that would trigger an ad for ‘I Like Turtles’. Each of these actions can increase the effective cost of bidding for the campaign and make it harder to appear for one of the higher converting search strings.

Increasing bids around peak buying times can place ads in front of people more likely to buy for certain products. Any increase in return on traffic can increase the amount spent per click. The same applies for more efficient landing pages and sites, and effective targeting geographically and by device. The more likely a sale, the more that can be spent per click.

This activity within a more general campaign can have an effect on how effective certain query clusters may seem. It can also reduce a campaign’s visibility for certain terms. A Search Query Performance report does not tell the full story of what is happening with your campaign and the targeted terms, but it can hint at competitive activity.