Two Golden Keys To Paid Search Success: Scale And Efficiency
There are two key factors crucial to assuring the success of any large SEM optimization process: First, tuning the campaign so that it is efficient and second, scaling the account while maintaining the efficiency you've gained. Your key goal should be to identify incremental opportunity without losing profitability.
There are two key factors crucial to assuring the success of any large SEM optimization process: First, tuning the campaign so that it is efficient and second, scaling the account while maintaining the efficiency you’ve gained. In any SEM account there will be a relatively small set of keywords that are naturally efficient (usually but not limited to branded terms) but that only drive limited volume. As a result, the goal is to use the tools available to identify incremental opportunity without losing profitability. In my work, I use three processes to identify qualified incremental opportunity: keyword expansion, impression share analyses and sophisticated bid management logic. When used together, these three processes can effectively identify opportunities for growth while keeping ROI/CPA under control.
The most basic approach to identifying incremental opportunity is through keyword expansion. Blindly expanding keyword sets has the significant risk of being unprofitable. However, if you have to expand blindly, be sure to implement a robust negative keyword set—limiting the potential detrimental impact of new, unproven terms. A more efficient methodology is to use the AdWords search query report to identify queries that have been broad or phrase matched by terms that are currently in your account. Adding terms which have converted (assuming the account is using Google conversion tracking) in exact and broad match pre-qualifies their relevance, mitigating risk as these terms are likely to continue to convert. While this process is not as fast, risk is negligible.
If you’ve had a meeting with a Google account team in the last 6 months or listened in on any holiday webinar, you’ve heard of impression share—a fantastic tool for identifying incremental traffic opportunities. Impression share (IS) metrics are available at the campaign level and consist of four unique data points:
- Impression share: the percentage of impressions garnered relative to total impressions available
- Lost IS (rank): the percentage of lost impressions due to rank
- Lost IS (budget) the percentage of impressions lost due to budget limitations
- Exact match IS: the percentage of exact match impressions garnered relative to total exact match impressions available.
If you take the sum of impression share, lost IS (rank) and lost IS (budget), the total will be 100%. These indicators are directional in the sense that they do not give truly actionable insight outside of whether you need to increase rank or budgets to get more impressions. That said, if you’re using Google conversion tracking or have the ability to track revenue/conversion data at a campaign level, it’s easy to see which campaigns are converting and have incremental volume opportunity.
This report starts to get really useful when you add an additional column, Ad distribution: with search partners. In general, Google.com tends to deliver more qualified traffic than its syndicated search partners, so with distribution-level insight at the campaign level, you can determine the average rank necessary to maximize Google.com traffic without maximizing syndicated traffic. Analyzing keyword level data to identify terms that are performing well with a rank that is above (the integer value is greater than) the “ideal” rank will provide an actionable data set to work with.
On the topic of ranking, before maximizing volume by pushing terms to a specific rank, it’s worth running a rank-based analysis to identify the ideal rank in terms of performance. Holding quality score, ad copy and landing pages constant, there is an “ideal” rank range where terms will maximize CTR relative to CPC (for more insight on ad rank calculations, see my post, how to get more PPC traffic for less money). In other words, analyzing the relationships between rank vs. CTR and rank vs CPC will yield a range (for example, 2.5 to 3.5) which has the most lucrative ratio of CTR to CPC. The image below shows an example using branded data:
When determining the ideal range, be sure to factor in conversion data to identify the max CPC you are willing to pay (through substitution and knowledge of the AdWords ad rank algorithm, we can determine: Max bid = goal ROI * average order value * conversion rate). When looking to maximize volume, aim for a rank that is at the more aggressive end of the ideal rank range. As this is simply a rank-based analysis, conversion rate is not expected to change as rank changes. If the goal is to boost conversion rates, other optimization techniques such as pre-qualifying traffic via ad copy and landing page tests are more likely to help.
The third tool for scaling with efficiency is bid management. It’s easy to lose track of ROI when managing terms to a given rank. This is why it’s important to identify a range within which you can manage for efficiency. Adjust bids to keep average rank at the keyword or match type in the ideal range (for example between ranks 2 and 7 to ensure first page exposure), then you can increase bids when ROI is above goal and decrease bids when ROI is below goal. This task can be tedious, which is why AdWords has developed conversion optimizers and rank-based optimizers (however both tools cannot be used in tandem). This is why most SEM software providers will allow logical rules to be implemented at the keyword or even match type level to automate the process, respecting the rank and efficiency parameters.
There’s a lot to think about when trying to efficiently scale an SEM effort. As always, when using each of these tools, often times together, don’t lose track of the underlying goals of the account (ROI, CPA, CPC, etc) as a byproduct of driving incremental volume.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.