There are a multitude of factors that determine success in search marketing ranging from intangibles (such as advertiser quality) to data driven dynamic strategy development. One can spend thousands of hours micro-managing campaigns only to realize the more work you do, the more work there is to be done. Don’t you wish there was a cheat code? A game genie for SEM? Unfortunately, there isn’t a set-it and forget tool that has proven to function perfectly, so until that’s been developed, we have to settle for best practices with the tools we’ve got at our disposal.
In the first two posts in this series, I covered methodologies behind a match type analysis and prioritizing efforts via top keywords. In this post, I’m going take a step back to discuss organizing an account, the ramifications on relevance and in turn, quality score.
Account organization is the first and most important step in the optimization process—it’s the foundation for success. A well organized account allows effective and efficient management of an SEM effort, freeing up resources that would otherwise be bogged down in the ongoing maintenance of the account. So, what is an organized account? And what differentiates an organized account from an unorganized account? These are good questions and I’ve devised a short questionnaire below to determine if your account is well organized.
At the campaign level:
- Are brand/branded terms isolated from non-branded terms?
- Is content match isolated into unique campaign(s)?
- Are site placements isolated into unique campaigns?
- Is there a unique campaign for each product/product line?
At the ad group level:
- Is there visibility into top performing keywords?
- Is each ad group focused on a unique product?
- Does associated creative speak to the product?
At the keyword level:
- Do all terms in a given ad group have a similar quality score?
If you answered “no” to any of the questions above, then there is work to be done. You may have caught onto the theme of my questions: more organization is better than less. That said, there’s a fine line between being well organized and going overboard. Furthermore, not all advertisers are the same, so to understand the logic of organization, we must think about advertiser and search engine limitations as well as the tools available for optimization.
For each advertiser, there are unique limitations. Some advertisers represent one store whereas other advertisers act as an affiliate/agency representing multiple stores and/or product lines (as an example, think of the online Apple store versus Best Buy, or The North Face versus REI). Representing one store is definitely the more simplistic model as there is only one brand to manage. Nonetheless, in both cases, all brand/branded terms should be isolated into a unique campaign. If representing several advertisers, then simply breaking brands into unique ad groups should do the trick. However, if there are too many terms (more than 30 or 40 per brand—this is a guesstimate of a maximum number of terms per ad group based on the logic that it’s difficult to have more than 40 highly related keywords) then it may be necessary to make a unique campaign per brand.
Because quality score is calculated at every possible level (keyword, ad group, campaign and account), it’s important to try to make each level as organized and relevant as possible. So, if you do represent multiple brands, then having unique accounts for each property/brand will boost the overall quality score of each account as relevance across the account will increase. If you are a more significant spender and have the benefit of support from the search engines directly, then request a master my client center account (MCC) from the AdWords team and have them generate unique accounts for each of your brands/properties. An MCC provides the convenience of one consolidated account (dashboard and reporting) with the benefits of unique accounts.
Quality score is determined from the relationship between the exact match of a term (the term itself), the ad copy associated with it, the landing page and historical CTR (assuming there is history). As a result, all match type iterations of a given term (exact, phrase, and broad) can live in the same ad group because the quality score will be the same. The only match types that should be isolated (for both dashboard and optimization purposes) are content match and site specific placements. All content match terms should be in unique campaigns; as content key performance indicators (KPIs) dilute campaign and ad group level data so significantly that it becomes almost useless. Similar to content, site placements are unique and optimizations should not be made in conjunction with search.
In terms of campaign level organization, each product or product group should have a unique campaign (depending on size constraints). Within each product based campaign, ad groups should be organized by type: navigational, informational, and transactional. Top performing keywords should be isolated into unique ad groups for enhanced control. Ad copy should be relevant at the ad group level and specifically tailored to the user’s stage in the buy cycle (informational, navigational, transactional). If possible, use product specific landing pages that contain user-friendly information regarding the product as well as a simple click path to purchase. If all of these processes are in place, then quality score should be consistent within each ad group.
While everything I have mentioned is designed to help in the long run, there is a significant caveat: overhauling account structure is a time consuming process and if not executed properly, quality score will be negatively affected. For this reason, whenever we overhaul an account at our agency, we start with top keyword isolation and ad group level changes as to maintain the keyword/ad copy/landing page relationships, thereby preserving quality score on the highest volume terms. Mirroring campaigns and overhauling [the newly mirrored campaign] is also an effective best practice as it’s easy to revert in case the short term negative effects on QS are significant.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.