BLOG
« Blog Home

Make your ads geo-aware

Mar 19, 2014

 

Part One of a Series

The Dynamic Ad Playbook is intended to educate advertisers and agencies about the benefits of dynamic / personalized ad creative, how it works, and how it can be implemented in campaigns.

Did you know that dynamic display ads can be geo-aware? In other words, it’s possible for an ad to determine where the viewer is and update the ad content accordingly.

For certain campaigns, geo-awareness can substantially improve ad relevance, which in turn helps lift clickthroughs and conversions. Here are a few examples of industries where geo-awareness can add value:

  • Automotive (car dealerships)
  • Supermarkets and pharmacies
  • Retail/chain stores
  • Daily deals (or any advertiser that’s aggregating local merchants)
  • Travel (detecting a traveler’s departure location)
  • Event and concert tickets

Four ways to implement geo-aware dynamic ads

We’ll look at four approaches to implementing geo-aware dynamic ad creative. There are other approaches and variations not covered here. Just contact us with any questions.

1. Location detection via the ad server or DSP

Many ad servers and DSPs (e.g., AppNexus) are able to pass viewer location data into the ad creative. This is usually accomplished using ad macros that are inserted into the dynamic ad tag. Examples of location data might be City, State, Postal Code, or Latitude + Longitude.

In many cases, locations are inferred from the viewer's IP address. For mobile ad inventory, latitude & longitude may be available.

2. Location detection via the ad itself

When location data isn’t available from the ad server or DSP, a dynamic display ad can, in a fraction of a second, detect the IP address of the viewer and determine the viewer’s physical location using any number of lookup services or datasets.

In terms of functionality, performance, and outward appearance, this approach is no different than Example 1. It's only when you peek under the hood that things look a little different.

DON’T "SCRAPE" IP ADDRESSES

Using an IP address once within the context of a single ad impression is fine, but using a dynamic ad to "scrape" and stockpile viewer IP addresses violates many ad industry privacy, publisher, and ad exchange rules.

This is one reason why it may be better to rely on Example 1 wherever possible, since location detection is being managed by the DSP or server, in accordance with applicable ad serving and privacy policies.

3. Identify visitor location contextually

Let’s say a retail chain has 1,000 U.S. locations and wants to retarget website visitors with location-personalized ads, perhaps containing the address of the nearest store.

When visitors drill down to a location-specific page of the website (e.g., the store hours page for a specific store), the retailer can fire a dynamic ads tracking pixel and pass a store location code into the pixel, thereby associating that viewer with a specific store.

Here’s code for a hypothetical pixel:


Then later, when that viewer sees ads, the ads will remember the location(s) the viewer looked at, and that information can be used in the ad.

There are two major advantages of this kind of active approach versus a passive location-detection approach:

  • The ad isn’t relying on approximate locations inferred from IP addresses. The ad is using locations that the viewer actively expressed interest in.
  • This kind of first-party site data can be stored and later used by the advertiser for any number of marketing purposes. And if the advertiser synchronizes customer ID codes with the dynamic ad platform, then the dynamic ads will be able to leverage additional first party data that wouldn't otherwise be known to the ad.

The major disadvantages of this approach are:

  • Only those visitors who drill down to a location-specific page are captured. This is likely to be a small subset of overall visitors, but on the upside, these are highly engaged visitors, so they're probably "worth" more in terms of retargeting budget.
  • This approach can’t be applied to anyone who hasn't yet visited the advertiser’s website, so it can’t be used for new customer acquisition campaigns, whereas Examples 1 and 2 can be used for new customer acquisition.

BONUS AUDIENCE SEGMENTATION TACTIC

Now that you've implemented a dynamic ads tracking pixel that only fires for visitors who reach location-specific pages in your site, you should probably create a new audience segment in your DSP. You can do this by firing a separate segment pixel along with the dynamic ad tracking pixel.

Here are three reasons to consider using this tactic:

  • FIRST: these visitors are actively looking for store locations, and are therefore in-market for your goods and services, and therefore have a significantly higher Expected Value than visitors who, say, spent 10 seconds on your homepage and bounced. It will be extremely useful to be able to report on these groups separately.
  • SECOND: if you believe that different visitor behaviors have different Expected Values, you'll want to optimize your retargeting budgets and ad frequency accordingly. Perhaps people who visited location-specific pages are worth $6.00 CPM and convert best when retargeting impressions are front-loaded in the few days after they leave, whereas people who only look at one or two pages are only worth $3.00 CPM and can be retargeted more evenly over 30 days or so.
  • THIRD: you'll be able to upload the ad creative that makes the most sense for each segment. For homepage-bouncers, you can run non-dynamic brand-building ads, and for location-specific visitors, you can run location-personalized dynamic ad creative.

4. Capture visitor location via signup forms

This approach is similar to Example 3. If an advertiser has a sign-up page with a location field (e.g., city + state, ZIP code, phone with area code), then visitor location can be captured at sign-up and later used in retargeting ads.

One advantage of this approach is the higher granularity of data potentially available for ad personalization. With multiple location factors available (e.g., city + state, ZIP code, area code), a dynamic ad can make more informed choices about which message is best suited to the viewer.

DON'T TRACK P.I.I.

You don’t want to track Personally Identifiable Information (PII) for use in dynamic display ads. PII would include things like email addresses, phone numbers, postal addresses, names, etc. There's usually not much use for PII in ads anyway, but make sure you implement the dynamic ad tracking tag so that it's only receiving non-PII data such as area code, postal code, city + state, etc.

You read this far, let's chat

Contact us and let's talk about programmatic ad creative.

Contact us to learn more about the Canned Banners Dynamic Ads platform.

CONTACT US