By Caitlin Halpert, Sr. Director of Client Services
You know all about Alpha Beta, right? (If not, you can download the introductory guide here.) The basic idea is to identify your top-performing queries (Alphas), isolate them for total control of optimization, and carve out a broader testing ground for potential top performers (Betas) to emerge.
That said, there are distinct best practices to follow when you're actually building out the campaigns. Let's jump in (we'll follow up with a later post on maintenance and optimization).
Identify Alphas from Account History
The most difficult part of identifying Alphas from historical performance is to determine what performance criteria to use while making the determination. These criteria are up to your team; we recommend a combination of conversion volume and/or click and impression volume. Once you've determined what those criteria are, you can then apply filters to either AdWords or Excel to identify the queries that meet them.
Use Account History Correctly
Account history has the potential to be either helpful or misleading when building out new campaigns. You want to be cautious while analyzing it to ensure that you don’t accidentally draw misleading conclusions.
Identifying top-volume queries is one way that you can use historical performance to guide you in creating new campaigns. It’s a safe bet that you should consider adding any keywords or ads that have been successful historically. Keep a running list of queries (and their corresponding performance metrics) that you want to consider. Those words that meet your team’s criteria for becoming Alphas will get added.
Before using historical data to conclude that a word is an underperformer, you should carefully audit the account structure for proof that it had a fair chance in the first place. It would be easy for a keyword with good potential to have bad performance metrics if it lacked negative keywords, had bad ad copy, or had any other number of disadvantages related to poor account structure.
Lastly, remember that historical data will not give you insights into omitted keywords.You always want to supplement your analysis of historical performance with research of your own.
Be Careful Using Tools to Identify Alphas
The Google Keyword Planner tool can be used in some circumstances, but it's not required, or even necessarily recommended, for identifying Alpha keywords (it is absolutely recommended for identifying new keywords, just not breaking them out into Alphas). The disadvantage of using this tool to identify Alphas is that it only includes estimates for volume and does not have any basis for providing performance estimates.
You might use this method if you don’t have reliable performance history, or if you have a client who is willing to take risks. Be warned, though, that this method can create high volume without corresponding high performance.
Identify Betas
It is often easiest to build Beta terms from a client account's site structure, the Keyword Planner, and other research tools. To narrow Beta terms, run your proposed list through Google’s keyword planner to ensure each keyword picks up a decent number of impressions per month. Again, the threshold will vary, but you generally don’t want to add keywords with fewer than 100 impressions per month. Reduce redundancy and only add more variations of a keyword if there is enough volume to support a custom headline. Each beta should be meaningfully different. Here are some tips:
Remove articles
Don’t include word order variations
Don’t include singular/plural variations
For some accounts, you can consider adding Gamma campaigns to the mix. These are campaigns that are set to Broad match, in accounts that have chosen to implement Broad Match Modified for their Beta campaigns. Gamma campaigns should have very few keywords to limit cross-matching across terms.
Set Bid Prices
Bid prices are generally set based on historical data. (The strategies employed will vary.) Bids should be set based on a combination of target efficiency goals, historical bids, and Google's recommendations. We recommend ensuring that your Alpha keywords have higher bids than their corresponding Beta keywords. Once you push your campaigns, there will be more data to adjust bids, so initial levels are not critical.
Use Broad or BMM for Betas
It is generally best practice to use BMMs for your beta campaigns, then test pure broad (gamma) campaigns at a later date. At 3Q, we will create new campaigns in either match type when our legacy campaigns are doing well and are not on automated bidding. We gradually transition to our new campaigns to ensure that we don’t disrupt existing performance and to ensure that we collect enough data for Google to run tCPA Smart Bidding. Once our new campaigns are built, they run parallel with the existing campaigns. This will allow them to gradually take traffic from the legacy campaigns until the majority of traffic drives to the new campaigns.
Consider Augmentations
There are several popular additions to Alpha and Beta campaigns. Some teams will build out dedicated campaigns for Competitors or ‘Exec’ terms (specific keywords the client has certain goals for). These are not standard, but they are common.
Gamma campaigns are another popular addition to the basic structure. These are campaigns made up of Broad match keywords that are designed to push additional traffic beyond your Beta campaigns (when the Beta campaigns are using BMM). These are appropriate for clients who have higher budgets and are interested in achieving greater reach.
Use Mirror and Bleeder Negatives
One of the last things that you’ll want to build out are the negative keywords that your account will require. This includes replicating your Alpha keywords as mirror negatives to our Beta campaigns, as well as creating lists of underperforming queries from your historical campaigns and adding them as regular negative keywords. All of these negatives should be added in Shared Negative Lists that can be easily maintained at the account level.
In addition to all of the negatives, you should include extra negatives to block any other queries that you aren’t interested in showing ads for. This could include competitor terms or products you don't sell.
You should maintain Shared Negative Lists for the following categories of terms:
Alpha negatives: exact-match negatives applied to all non-alpha campaigns
Irrelevant negatives: includes adult terms and bleeders
Performance negatives: keywords you’d want to retest in the future if conditions changed; old non-converting alphas
Cross-matching negatives: ad group-level negatives for improved CTR; check periodically for cross matching of queries or patterns of queries
Ad Templates and Customization
Creating new ads can be time-consuming; employing a template can save time. We recommend analyzing the historic performance of different ads and creating two lists: one of the types of ads, and another for the attributes that the ads have.
In this case, you can create a template that substitutes a product name for “Widgets” and allows you to create ads quickly. You can increase the effectiveness of the template by analyzing the performance of different ad attributes before deciding what value prop and CTA to use by filtering an ad report to look at performance of ads that contain certain lines.
For example, you might notice that ads with “Order Now” and “Free Shipping” do really well, but ads with “Save” aren’t so hot. You can use that data when constructing templates for products, categories, or any other part of the account. Once you've got your template built, you can substitute the relevant product, or category names into the template to create custom ads for each SKAG. After finishing your ads, be sure to consider which ad extensions are relevant for your client.
Ad Rotation Settings
Always ensure that your ad rotation is set to "Optimize" before you push anything live.
Follow those practices, and you'll be able to build yourself a tidy, ROI-friendly Alpha Beta campaign. Again, stay tuned for how to maintain and optimize your campaigns to keep them as profitable as possible.
