The hardest step to take is the first one. PPC account analysis is the same way. A request for an account analysis sends me into analysis paralysis.
The necessity of an account deep dive can manifest for a variety of reasons. Before you succumb to analysis paralysis, read through these simple account analysis steps to help get the analysis ball rolling.
Review Account Structure
The first step in an account analysis is to understand the account structure. This is the bedrock on which rests expansions, testing, and further analysis. Account structure can mean a lot of things, so here is a quick list to get you started:
- How many campaigns are in the account? Ad groups per campaign? Keywords per ad group?
- Are the ad groups segmented by matchtype?
- Which bidding method is being used?
- Check the campaign settings.
- Are any campaigns limited by budget?
- Is there a uniform naming convention?
- Maintaining a consistent naming convention throughout the account will help keep the account organized and will help with filtering.
Make sure to check out the device performance. Mobile performs different from desktop which performs different from tablet. In an account I recently analyzed, I found that over half of the clicks were from mobile, but these mobile clicks only contributed to 29% of the total conversions.
Device performance can lead to deeper analysis. In some cases, you can simply utilize device bid adjustments. However, taking a deeper dive into the mobile user experience will help you better understand the data.
While demographics reports may not yield big “Aha” moments, they do increase your understanding of the audience. For example, knowing which age range converts more will help you write ad copy that resonates. You can find demographics data in both Google Analytics and Google Adwords. I have found demographic reports to be very helpful when optimizing my display campaigns. Analyzing which age or gender does not perform well can lead to a quick, cost-efficient bid change.
Time Of Day Analysis
It is always helpful to run a time of day analysis, but especially helpful if you have any campaigns that are limited by budget. If your campaigns are running during hours of the day where they are receiving no conversions, by bidding down or turning your campaign off during that time period you can reallocate, what would have been wasted spend, into campaigns that are converting. I have found it helpful to look at a large chunk of time, as large as the last year. This will help smooth out seasonality or performance fluctuations. The large chunk of time will reveal areas with zero conversions. Hours of the day, in the last 365 days, with zero conversions are a safe bet to either bid down significantly or turn the campaign off during those hours completely. You can always run these changes as a test.
Search Query Analysis
Developing a rhythm of search query analysis is a good habit. I have had campaigns with great matchtype segmentation and a healthy list of negatives and have an ad group mysteriously ramp up spend only to find that one of the keywords started triggering random queries. For example, I have an education client whose business operates under a lot of seasonality. This being the case, August and September are hot months. There will inevitably be some random search queries triggered by keywords.
When reviewing the search query report, it is also helpful to look at which queries have the highest cpa. While they may have converted, the cost per conversions could make those terms unprofitable.
If you are performing an account deep dive, the search query analysis can also serve as a guiding light toward tighter account organization. If there are no shared negative lists in the account, the search query analysis will reveal common queries that can be excluded across multiple accounts. Utilizing the shared negative lists, in the Shared Library, will save a lot of work in the future. Instead of combing through lists of ad group negatives you can troubleshoot from one list that has multiple touchpoints in the account.
To help get you thinking about where to start with a shared negative list here is an example. I have a client who sells products that are higher priced in the industry. Because we know we cannot compete on price we created a shared negative list for words like; “discount,” “cheap,” and “inexpensive”.
Expanded Text Ad Review
Even though the transition to expanded text ads happened a while ago, there are still accounts running ad groups with no expanded text ads or only one running one or two expanded text ads. So, first things first, make sure that all of the ad groups have included expanded text ads. These days, Google recommends having at least three expanded text ads in an ad group.
Another thing to check are the ad rotation settings. Are the ads set to optimize or to rotate indefinitely? Whichever ad rotation you choose, be sure to set an ad copy testing plan moving forward. Testing ad copy is still possible with the optimize rotation settings, it just requires the use of Google’s experiments and drafts.
Ad Extension Analysis
Ad extensions are a quick way to improve click thru rate and make your ads look full and pretty on the SERP. Once you have ensured that all available extensions are in the account, you can review performance. Are there any extensions that are not performing well? Remember that promotion extensions are available, but only in the new Adwords interface.
There you have it, a quick and simple start to performing an account analysis. The idea is that each of these analyses will allow you to expand your understanding of the account and to begin building a well-informed account strategy. One thing to keep in mind as you are running these analyses is that there is much more nuance than I covered here. For example, I have a client that advertises in three different countries using display, shopping and search campaigns. It would not work to run a blanket analysis and make adjustments as if all locations and all campaign types performed the same. A good place to start would be running a high-level analysis and then segmenting by location and then by either shopping, search, or display.
As always, let the data guide you and happy analyzing!