Geotargeting Challenges and Solutions for Mobile Banner Advertising

The age of the smartphone is upon us, where now more than half of all mobile users 18-34 have a phone that has advanced features including internet access and some kind of higher-resolution screen. Smartphones are essentially phones packaged with computers, and many have services that include GPS or some other kind of location-determining feature that can pinpoint a user to at least the city or metro level. With the proliferation of iPhones and Android-driven phones in the marketplace for key age group demos, there is a strong demand from advertisers to target users via these services and serve them banner advertisements (usually 320x50) that are relevant to their immediate location. When you target someone based on their immediate location, it is known as Geotargeting.

However, there is a catch, currently, technology has not advanced to a point that most commercial ad servers can locate where a user is reliably without the user actually inputting their location manually. Since most users will not voluntarily give their location just to be served an ad, there is a high chance that location targeting done the traditional way that most advertising servers have done in the past, e.g. IP targeting, will be correct. In fact, many mobile data services tunnel their access in ways that do not consistently give the actual location of the user, and could be off as far as even being in the wrong state. This is in contrast to IP targeting desktop users, where IP targeting has been honed to a point that it is fairly reliable -- in other words, that a user whose IP says they are in Japan is in fact sitting in Japan.

The solution is to have the ad server write the location of the page (if relevant) or the location of where the user is registered (if the site requires or uses registrations) into the ad tag itself. This is called many things in different ad servers, but in DFP, this is called key-value targeting. So, if you have a page that is for a restaurant in Las Vegas, write "metro=lasvegas" into the ad tag. If you have a site that requires or cookies users who are registered, and they put their home location as New York, write that into the tag. If the user did give their location when they signed into the app or mobile web site -- write that location. This can then be targeted instead of trying to sniff by IP.

Will this work in all cases? No. But it is a best-effort implementation for an issue that currently limits the entire advertising industry. Future upgrades of various ad server products may fix this, but for now, this is the best that can be done given the unique limitations. This allows publishers to have some kind of reliability around Geotargeting in their mobile inventory.

If you are a publisher, your solution is above. If you are an advertiser, ask if the site or property you are advertising on has the solution above, if Geotargeting is one of the criteria you require or seek. If not, you have a ready-made solution you can suggest.

Read More

Decrease Your Ad Revenue with The New DFP Ad Tags (Google Publisher Tags)

What is DFP Ad ? DoubleClick's DART for Publishers (DFP) solution provides the scalability, reliability, and the critical tools to accelerate your digital ad operation.

After hearing all the exciting news and new features of the new DFP advertising tags (Google Publisher Tags), how could a publisher not implement them? After implementation of the asynchronous ad tags, the ads are supposed to load faster from single page request ad serving, you're supposed to be able to serve ads in emails and newsletters and DFP can serve passback ad tags that point back to DFP to choose the next ad.

That all sounds great, right? Well, we fell for it too. We implemented the new tags for several clients and encountered a few problems:

1. Ad Exchange CPM's Dropped:

The exact day we implemented the Google publisher tags, we saw about a large drop in CPM. The next day we saw close to 20% drops in CPM. Mind you, this was in Q4 when CPM's are supposed to consistently climb as we get closer to Christmas. This has happened every year in the past, the only difference here was the implementation of the new DFP tags.

2. Advertising Fill Coverage Dropped:

As soon as we implemented the Google publisher tags, we saw drops in ad impressions immediately. We compared this to the site's Analytics and saw no consistency. The traffic remained constant but the ad impressions dropped by an average of 10%. Therefore, the new DFP ad tags served a lower percentage of functional ads versus the old tags.

3. Non-existent Third Party Advertising Network Passback Capability:

Google claimed the new DFP ad tags enabled the capability to send passback ad tags to third party ad networks that would send any passback ad impression back to DFP to choose the next advertisement to show. This is a feature that DFP Premium has and is the optimal passback strategy for any publisher. However, this feature is non-existent for the Google publisher tags. Many DFP reps that we've dealt with have admitted this and said they will change the language about this feature.

Overall, the new DFP advertising tags have been a huge and expensive disappointment. We have reverted back to the old DFP advertising tags and are in the process with the other clients. We believe the drop in Ad Exchange CPM came from the fact that the new DFP advertising tags are iFrames. Expandable ads are not able to be served in iFrames without additional external code setup. Expandable ads go for high CPM's and the new DFP advertising tags aren't able to serve them properly which hit the Ad Exchange CPM's hard. The new iFrame ad tags could be the reason why the fill of functional ads dropped by an average of 10%. Expandable ads and certain rich media advertisements in third party ad networks weren't able to serve properly. The lack of a third party ad network capability was negligent communication on Google's part. It is important to ask your Google rep about these specific features first.

Read More