Managing Huge PPC Campaigns? Try These Best Practices
Reading search marketing advice, you will find an abundance of information about best-practices that are easy to apply to a campaign or ad group level. Common best practices such as keeping ad-groups small and targeted, with relevant and keyword specific ad-copy have demonstrated great results. Moreover, the value of managing match types and negatives to control or avoid poorly targeted impressions can’t be stressed enough. However, as anybody who has managed a large-scale search program understands, knowing the best practices is only part of the equation.
To be truly effective, you need to understand where to allocate your time applying these best practices. With potentially thousands of ad groups to choose from, figuring out where to focus your optimization efforts can be challenging. Evaluating your overall paid search program by business-relevant dimensions is a helpful way to identify this type of optimization opportunity.
Search campaign structure is typically optimized to meet the targeting requirements of the engines, but most marketers will find more value in analyzing performance of campaigns if they are segmented by characteristics relevant to their business. As you look at data through different lenses, such as business line, margin, geography, website or brand vs. non-brand terms you can get a better understanding of how performance is distributed and where the greatest opportunities lie. Starting at the highest level and then drilling down into specific ad groups allows you to pinpoint the best opportunities for optimization and growth.
To effectively analyze information at a category level, many of our clients find charting to be a valuable tool for identifying pockets of opportunity. I’ve provided an example of how this works using a standard “Efficiency vs. Exposure” graph below. In this three-dimensional bubble graph, we are charting the efficiency of a keyword set versus the average position and volume, for a fictional sportswear retailer. In the chart, average position is on the y-axis; the x-axis represents efficiency (as measured by average cost-per-acquisition) and the size of the bubble represents the total volume of conversions for each category of keywords. For example, the group of “brand” terms has an average position of 1.1, an average CPA of $5.05, and drove the highest volume of conversions.
By breaking up this chart into four quadrants, we can identify the key areas for improvement and prioritize our optimization efforts accordingly. More importantly, stepping through each quadrant in the chart can help us determine what types of optimizations we should apply.
Quadrant I: Stretch For More Volume
The first quadrant houses keyword sets with low cost and high position. These terms are extremely cost-effective, but with a higher average position than many terms it’s possible that they aren’t getting the exposure that they deserve. With these keyword sets, you will want to drill in and understand how many of your terms are already in position one, two and three, versus the number that are in lower positions on the results page. If you can increase your bids on terms in lower positions while still hitting your overall margin targets, you can drive higher revenues at an effective cost of acquisition. Many of your keywords may simply not be showing on the first page of results. In this case stretching to meet the minimum bid can be a simple way to increase performance.
Quadrant II: Consider Cuts
The keyword sets in quadrant II represent the groups most in need of improvement. These ad groups have the highest average position and highest cost, meaning they are the least efficient. Because the size of the bubble shows the number of conversions each group is driving, the largest bubbles represents your biggest opportunity. In the example above, the “golf” terms should be top priority. Now that you have identified this opportunity, you can analyze where the issues lie. Are there campaigns or groups that you could eliminate from your program? Eliminating inefficient spend allows you to reallocate dollars to higher performing programs. Are quality scores or conversion rates too low? Depending on the situation, you could focus on adding negative keywords or testing your creative.
Quadrant III: Focus On Quality
The keyword sets in quadrant III are less efficient than others, and yet you are paying for these terms to be in top positions on the page. As you did with your high-cost, high-position keywords in quadrant II, you can prioritize your optimization efforts around the biggest bubbles in the quadrant, addressing the high-volume terms first. In this quadrant, a poor quality score or conversion rate is likely to be the reason for low efficiency. Adding negatives to these groups can help refine traffic which should both improve the likelihood of conversion as well as increase quality rankings. Similarly, evaluating raw queries and adding phrase and exact match terms to groups can help shape your traffic and improve performance. In addition to refining match-types, testing your creative copy will help lift conversion rates and overall efficiency.
Quadrant IV: Expand, Expand, Expand
Congratulations! The keywords in quadrant IV are your star performers. Hopefully, as in our example, your biggest bubbles appear in this low-cost, low-position quadrant. You should not need to spend much time optimizing the structure of these campaigns, but what time you do spend is most valuably focused on expansion. Try using search query data or popular keyword research tools to find new, profitable terms that allow you to drive more revenue from these groups.
While most marketers have an intuitive idea of where the areas of opportunities lie in their paid-search programs, using a graphical approach to analyzing, prioritizing, and optimizing your programs will make your process more scientific and repeatable. As you identify opportunities and iterate by drilling down on the most out-of-line campaigns, you should begin to see significant improvements in overall return on investment.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.