Everything you need to know about SEO, delivered every Thursday.
Busted: Three Boneheaded Paid Search Myths
Paid search marketing, with its complex bid management techniques and daily campaign management challenges, is difficult enough. And it doesn’t help that the industry’s thousands of vendors and consultants each think they have the inside secrets on how to optimize campaigns—and aren’t afraid to share their “advice” with you at every turn. This constant barrage of input combined with the need to test everything, understand a dizzying amount of data and build a workable strategy for bid management can make many search marketers nervous they’re not maximizing ROI from their paid search programs.
There are three myths that seem to still be circulating in the industry about the complexity and difficulty of paid search optimization. It’s time these myths got busted so you can focus on the real optimization techniques that matter.
Myth #1: Google will pick the best creative for you
Most marketers think if they develop a campaign on Google, the search engine will automatically determine and serve the best-performing ad copy from the available set. Many marketers believe this means they should continually add creative, since the engine will take care of testing for you. But all that testing may not be helping as much as you think. Google determines which ad copy to serve based on which version has the highest click-through rate, but your highest CTR ads aren’t necessarily your best-performing ones. To get around this problem, you need to do two things: set the creative rotation in Google to “show evenly” to perform an accurate test, and then do your own tracking and measuring of ad copy to see which ones are performing the best.
Start by determining what you want to test, and defining a start and end date for your test. Once you implement the ad copy, ensure you are tracking conversion at the creative level using the dynamic insertion parameters provided by the engines. Also, be sure to choose a relevant metric like conversions-per-impression, instead of just click-through rate. Compare each creative in an ad group vs. the average of the ad group without that creative. By doing this, you find out which creatives are driving the most conversions, and once you know this, you can jettison poor-performing creatives and manually rotate higher-performing ones into rotation.
The bottom line: Pause creatives that come in below the average; continue to monitor those that are near average for any changes; and keep running creatives that are out-performing the others using a metric such as conversions-per-impression rather than the relatively meaningless click-through rate metric.
Myth #2: You have to look at all prior clicks for accurate bidding
Most marketers bend over backwards trying to attribute conversions back to every click in the visitors’ path to come up with accurate bid prices. It’s understandable you’d want to accurately attribute conversion to each keyword by looking at past clicks, but multi-click attribution can be difficult to track and even more difficult to understand and make actionable. Many marketers avoid tackling the attribution problem head on by simply over-investing in early-consideration terms, even if the ROI isn’t there, because they believe these terms influence purchases down the line.
The truth is, you’ll get a reasonably accurate picture of conversion and ROI for campaigns by attributing value only to the last click, because the majority of people do not click on multiple search terms before making a purchase. In fact, research into a leading client of ours spanning hundreds of brand-name and product-name keywords found that only 26% of paid search conversions had more than one prior click on a paid search term. That means three-quarters of all paid search conversions happened after just one click—so if you measure back just one click, you’ll get a pretty accurate measure of ROI. What’s more, we found that almost half of these multiple-click conversions had the same term for all prior clicks—making last click attribution accurate in more than 85% of all cases.
The bottom line: Evaluating and implementing a complex scheme for attributing conversion across multiple clicks can be valuable, but the impact from a measurement standpoint is likely small. Instead, you may be able to get a lot more value from focusing on other low hanging fruit for optimization such as keyword expansion, adding negatives or creative testing.
Myth #3: “Portfolio management” is really, really hard
Many marketers get stuck believing that optimizing their keyword portfolio to increase profit is inordinately complex. As a result, they turn to outside help, and in many cases will simply hand over bidding to a third party. The truth is, at a basic level, portfolio optimization doesn’t have to be hard, and in many cases, it’s downright easy.
When thinking about portfolio management, it’s crucial to calculate a “headroom” metric—or the bid price minus the cost-per-click. For example, if you bid $2 on a keyword and your CPC ends up being $1.40 after the auction is complete, you have what is known as “60 cents of headroom.” You could pay 60 cents more for this keyword and still meet your pre-defined ROI targets.
Using this metric, you can look at all of your keywords in your portfolio that have substantial headroom. If these keywords are in position #1 in the auction, don’t change anything, as your bid is already working to place you at the top of the heap. These keywords are maximizing your profit, and generating excess returns for you. However, if you have keywords with headroom where you are in a competitive position in the auction—maybe 2nd, 3rd, or 4th place—it might pay to raise your bid on these keywords to move up in the rankings. You’re already getting several top position keywords for lower than you’re willing to pay, so it’s OK to pay a bit more for some other terms.
The bottom line: By increasing your bids on competitive keywords with headroom, you can wring more margin out your overall portfolio while still meeting your ROI goals.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.