Adwords analysis is the basis for all Adwords optimization. Adwords analysis is heavily based on reporting features. When starting Adwords analysis, there is one thing to keep in mind: What is the goal of the analysis? Are you trying to achieve lower cost? Is it to measure Adwords against objectives? Whatever the goal, it’s important to settle on one before beginning Adwords Analysis.
The next step is deciding what part of Adwords are you wanting to analyze. If you’re wishing to utilize keyword performance, there are several reports that can help you. It’s important to note that one of the best ways to measure your Adwords account is conversion tracking. If you don’t have conversion tracking installed, you’ll be unable to determine which of your ad groups or placements is producing results. Assuming you have that installed, keyword analysis is a breeze. Most of this can be done through the main keyword reporting tool. To do more advanced analysis such as search impression share, or performance by segment, you can use the filtering tool to filter and see only the keywords that are relevant to your search.
Another common type of analysis is mobile device performance. To isolate and measure how mobile devices are performing in Adwords is not a straightforward task. One method is to use the dimension tool report to identify aggregate performance. Another option is to enable segmentation in the main view to view device performance on an ad group or campaign level. This type of analysis can measure the performance of mobile devices and/or tablets against desktop devices for metrics such as cost per click, conversion rate, or cost per conversion.
Another type of analysis is Display Network optimization. When analyzing your Display Network campaign, there are multiple factors you will need to analyze. You’ll need to look at placements, demographics, and or display keywords and lists if you’re using those. Analysis is best performed with a benchmark, but you can also compare different targeting settings against each other. For example, you can very easily compare the demographic 18-25 with the demographic 50-65 from a cost and conversion level. This type of analysis is very effective in squeezing the ultimate amount of performance out of your Display Network campaigns.
A favorite of ours is historical performance analysis. This type of Adwords Analysis takes a segment, such as the last 30 days, and compares it to another segment ideally one year ago to have a full year-over-year analysis. This type of analysis is immune to seasonal fluctuations, but if things are drastically different in targeting or another perspective int hat past time frame, it can be valuable analysis. There is no way to do analysis with being able to fully control events and relevant data, but a well thought out Adwords analysis can provide key insights into changes that need to be made on the account.
A few other notes about Adwords analysis. It is very important to have a complete data sample when performing an Adwords analysis. This is important because you must have statistically enough information available to provide a meaningful conclusion. Here’s some example data sets which have enough information to provide a conclusion:
On a keyword level, 200-500 impressions can generally give you insight into the click through rate of the keyword. In general, 10-20 clicks can give you insight into the performance of the keyword. In an ad, about 50 clicks minimum is needed to validate whether or not a given variant of an ad has an appropriate click through rate and can be kept or discarded. Since campaigns contain a multiple of ad groups and keywords, there is no fixed benchmark on how much traffic must pass through it in order to determine a campaign’s success or failure. To further complicate things, under performing campaigns can be optimized by adding additional keywords and/or ad groups to expand the campaign in different ways as an example of implementation based on analysis.
Situations where you are definitely not going to have enough data to make a conclusion are within the first couple weeks of the campaigns launch, and if there are just a few clicks on any given keyword or ad group. Don’t be quick to write off the ad group in your analysis without fully examining the keyword click report. For example, if most of your clicks come from a specific keyword, don’t be hasty to write off the ad group as a whole. The other keywords in the ad group may not have gathered any traffic and all, but may still be effective keywords.
Another pitfall in Adwords analysis are search funnels. Search funnels are an advanced concept where a user will initiate a search on a broad term, such as “running shoes”, and then refine their term to add additional qualifying keywords to it, for example then searching for “Nike running shoes”, and then subsequently “buy Nike running shoes”, and then “lowest cost Nike running shoes”. The most frustrating part of Adwords analysis is it is often very difficult to determine if keywords are part of the upwards search funnel. Sometimes after doing a light or shallow Adwords analysis, it is possible to conclude that certain keywords or ad groups are not performing, and remove them. This can sometimes have the unintended effect of removing keywords which are part of the buyers search process. The net effect is that other keywords that were performing well suddenly stop mysteriously. It can be very frustrating to analyze an account, make changes, and then have keywords that were previously performing well stop. This can be possible if you’re not doing a proper and thorough analysis and evaluating all the variables. Since Adwords is very complicated, it can be easy to lose track of an individual detail or setting without keeping the whole context of the account in mind.
Let these guidelines help you do a proper Adwords analysis. And remember to always double or triple check your conclusions to make sure they are valid.
