At SMX East this morning, I sat in on the Improving Your Geographic Targeting Tactics On Google & Bing session, in large part because I’m far more familiar with SEO than with paid search, so I like to stretch myself and learn more about the advertising side of the local search equation. The session was moderated by Matt Van Wagner (President of Find Me Faster), and featured speakers Kevin Lee (CEO of Didit.com), Sam Owen (PPC & CRO Account Manager of Hanapin Marketing), and Paul Corkery (Senior Operations Program Manager at Microsoft).
I actually have quite a lot of familiarity with the background technologies used to accomplish geo-targeting (check out my article from 2007, “Geolocation: Core To The Local Space And Key To Click-Fraud Detection” if you want a tutorial on how the locations of internet users can be identified). So, while there were not necessarily any big surprises for me in this session in terms of how geotargeting of search advertising functions, I did learn about refinements in the features for setting up Google AdWords and Bing Ads campaigns that were entirely useful and downright clever. Knowing how the technology functions is still a far cry from knowing how to use it effectively, as I’ve found time and again when setting up and managing ads—the devil’s most definitely in the details!
There are a number of reasons why one would want to geotarget ads, even for products/companies where you don’t particularly feel that there is any obvious need or requirement to make an ad locally oriented. While perhaps the main reason one may geotarget ads is because one only offers services/products within specific areas, there is a huge and compelling demographic component involved with geography that can influence CTR and conversion rates as well. So, even if you think you don’t have a use for geographic targeting of ads, read on through some of the takeaways I’m highlighting from the SMX conference and revisit some of your assumptions! It may just be that almost all campaigns should be assessed for possible geographic segmentation opportunities.
One non-obvious tactic that both Kevin Lee and Sam Owen highlighted is to make sure to use geo-specific language in the ads and to have specific landing pages for geotargeted ads as well. I know from general PPC ad management best practices that it’s often ideal to include the keywords that are being targeted in both the ad copy and the landing page since there’s clearly a psychological reassurance or reinforcement when searcher’s see this—that what’s being advertised is indeed what they’re looking-for. It may also be that there’s a reticence that’s been trained into internet users over time because they’re barraged with ads unrelated to them or what they may be desiring, so increasing the signals around ads that convey “this is what I want” will improve performance.
Where geotargeted ads are concerned, the geographic area name is not always part of the keyword query, but including that name apparently increases the effectiveness of the ad and landing page.
One of the most strategically interesting reasons to use geographic targeting is the ability to bid on keywords by geography, allowing one to maximize click-through-rate (CTR). For example, you could adjust bids by geographic area to push exposure more in areas where you get better CTR. Or, remove geographic areas that just don’t perform well for you. Using this you could pump up your campaign’s overall CTR, keeping your bid costs lower, your ad rankings better, and your conversion rates optimized.
Both Kevin and Paul spoke of the issue of accuracy which has been a limitation of the technology for some years, so if you do geotargeting, it’s great to be aware that it’s not possible to perfectly limit ads to display only to people within the area you specify—a percentage of ads will still get displayed to people not in the right area. This doesn’t negate the effectiveness of geotargeted advertising, but it’s good to know the reality of the situation. As Kevin eloquently states, “Truth is that geotargeting just increases the odds of the person being where you want them to be.” Paul even went further to educate how the situation is: the more granular the geotargeting, the higher the percentage of error that creeps in. He perfectly illustrated it with graphics that showed an increasing section of uncertainty as one goes from country-level targeting to zip code targeting. When targeting ads by country, the error rate is only a fraction of a percentage, most likely, while ads targeted at the most granular level of ZIPs may be as high as one or two percentage points.
Kevin recommended using geography as a “Big Data” predictor for demographic and psychographic audience segments. Using geography, you could predict in advance whether people in certain locations have the income, age, wealth—even BMI/weight—necessary to predict interest in your products and therefor predict conversion rates as well. Better audience targeting delivers to you “bid leverage” with PPC—if you geotarget ads, you can leverage your improved CTR to get better bid rates, or increase your bids for segments where you know the conversions will pay for greater campaign budget spend. Kevin’s company, Didit, provides a map tool for helping segment geographically based upon: household income, family size, wealth, shopping cart size, conversion rates, and client-specific needs. Geography can be broken out further by other criteria, such as political affiliation, personas, real estate values, magazine readership, Apple/Mac/iPad penetration and other demographics/psychographics.
One of Sam’s case studies compared performance on a non-local-specific product that was similar to insurance. They ran nationwide, non-local ads and compared with geotargeted, local-specific ads and landing pages. While the difference wasn’t huge, there was indeed a difference. The localized ads achieved a 3.25% CTR versus a 2.99% CTR. In another case, by using geographic targeting instead of running ads nationwide, his company achieved a 14% increase in conversions over the previous year’s campaign.
Sam pointed out that prior to Google launching Enhanced Campaigns, geographic targeting was dauntingly more complex and difficult if you also broke campaigns out by mobile/tablet/desktops. For instance, running four campaigns, segmented by mobile/tablet/desktops and bid separately by fifty states would’ve required six hundred campaigns!
4 Ad Campaigns x 3 Segments x 50 States = 600 Campaigns
Sam believes that the Google enhanced campaigns may be a good thing, and the geographic management may be the best part of the system. He recommends only using geographically separate campaigns now that need geographically specific keyword bids.
Bing offers an “advanced location” option—it allows one to target people who’ve sort of telegraphed a “location intent”—these are people who Microsoft believes are more inclined to do business in your local area. Bing identifies these searchers in a few ways, such as when the user includes location terms in their search queries. Examples that Paul provided were:
Bing also uses some publisher signals and contextual signals such as if a searcher in America is searching on Bing.co.uk site, then they can infer that the user might be planning a trip to the UK and maybe looking to buy local products/services in that location.
Paul also provided a few other subtleties and tips around Bing which I thought were interesting:
After sitting in on this session, it’s clear that there are quite a few complexities involved with geotargeted advertising. I’ll bet that many companies are not effectively managing geotargeted ad campaigns. But, if you do, you can improve your ads’ performance, sometimes significantly.