LOCATION BASED PROGRAMMATIC ADVERTISING

Case Studies

Location Based Programmatic Display Advertising – an intro

It’s possible to serve ads on mobile phones only on/in specific locations. These locations are defined by geo-fences (circles) of as small as 25 meters (radius). Mobile phones can forward a GPS based location within an adrequest. This location can be as accurate as 10 meters.

In our geo-tool it’s also possible to (automatically) upload hundreds, thousands or even hundreds of thousands of addresses (based on zipcode and housenumber). We are not aware of any other geo-tool (within a DSP) that is able to process and actually serve on these huge amounts of addresses, within one campaign/line. The tool allows you to change the radius of all (!) these geo-fences in one time/with one action. It’s also possible to apply individual sizes for each geo-fence.

Problems and challenges

The functionality has some problems. For example; many of our clients ask us to serve a banner on their specific phone – by setting up a geo-fence above their company location. In many cases we are not able to show the client a specific banner. This situation significantly undermines the apparent abilities of serving of ads on specific locations. 

Some reasons for a person not seeing a banner:

• An active ad-blocker (in many cases a person is not aware of this and does not know how to disable it)

• Active ad-blocker based on app preferences

• The GPS location that the phone upholds does not ‘match’ the geo-fence

It is worth mentioning that although location transfer possibilities via browser (mobile web) are on the rise, a location transfer via an adrequest based on app traffic is - to date - still the most reliable method.

The fact that it is very hard to show a client a banner (based on a geo-fence ‘above’ his/her actual location) is problematic. You can even state the functionality (location based advertising) does not work. There are – however – possibilities to drastically enhance the possibilities of a client seeing an actual banner on his/her phone – when requested – and prove the technique works. For example, communicating the device ID of the phone in question enhances chances of seeing an ad (on that particular phone).

Luckily there is another method to prove the technique works. A method that we will describe in this case study.

The campaign

One of our clients (an agency) had an unusual campaign objective. Their client, a care provider for elderly people, urgently needed 18 new employees. The agency asked us what we could do in terms of Programmatic Display Advertising. Because of previous successes with the following campaign strategy, we recommended to serve ads ‘above’ actual locations of nursing homes and other ‘care’ related companies, government locations and hospitals. The client agreed, and we bought an address list containing 12.526 locations/addresses that match those criteria. Usually these kind of address lists are not hard to come by.

We created banners (standard IAB and a rich media interscroller) and uploaded the list into our platform/geo-tool. We set-up geo-fences ‘above’ all these locations and upheld a diameter of 110 meters. It’s worth noting that this campaign was the only marketing method active for finding these 18 new employees. Secondarily we launched a campaign with the exact same ads/banners, but without Location Targeting. Hence, the ads where served in the entire Netherlands. Overall, we served 1.428.000 impressions during a 3-week period. The good news; the client received enough applications to fill all the 18 vacancies. There is more to it though.

Results

• The main line (serving ads on these 12.526 specific addresses/locations) realized a 0,72% CTR based on the standard IAB banners. A 1,56% CTR was achieved with the Interscroller format.

• The landingrate for the IAB set was 74%, while the landingrate for the Interscroller was 56%.

As explained we launched a campaign with the exact same creatives, but without location targeting. These ads where essentially served in the entire Netherlands (again; no location targeting on specific addresses). The results were as follows:

• A CTR of 0,23% for the IAB banners. The interscroller achieved a 0,82% CTR.

• The landingrate for the IAB set was 44%. The landingrate for Interscroller was 36%.

What do these results tell us? Well, for starters that the banners served ‘above’ specific addresses/locations realized a significantly higher CTR. The same goes for the landingrate.

• The CTR of the standard IAB banners WITH location targeting is more than 3 times higher than the CTR of standard banners served WITHOUT location targeting.

• The CTR of the Interscroller WITH location targeting is almost 2 times higher than the CTR of standard banners served WITHOUT location targeting.

• The landingrate of the standard IAB banners WITH location targeting is approx. 85% higher than the landingrate of standard banners served WITHOUT location targeting.

• The landingrate of the Interscroller WITH location targeting is approx. 65% higher than the landingrate of standard banners served WITHOUT location targeting.

In turn, this tells us that the banners served WITH location targeting have a drastically higher CTR than the banners served without location targeting. Furthermore, the landingrate is higher within the Location Targeted campaign, while the landingrate within the campaign that is not targeted on any specific locations, is lower.

Apart from the actual filling of vacancies, these results prove that when ads/banners are being served on specific locations, and the apparent audiences present ‘in/on’ these locations have a higher interaction with the ads served (because they are relevant) – reaching audiences by serving on specific locations; works.

Reach out to learn more and/or to discuss what kind of Location Targeting options might help you reach your audience in better, more cost-efficient ways.

DEPLOYING A WHITELIST ON OPENRTB IN 2019

Here is why it’s extremely difficult to get traction. But, if you do - CTR’s above 40% - can be achieved.

We all have them; clients/advertisers that just want to serve ads on a specific whitelist. And right they are. Many reasons exist to only serve ads on a specific whitelist. The biggest ones? Brand Safety, reaching a very specific audience and distrust in actual Brand Safety tools. The latter because Brand Safety tools – for example - claim they will protect your brand/offer from being served on low quality publishers. Any expert knows that – if you want volume (regarding your campaign) – such a statement is unrealistic.

Let’s take a step back. Because literally every hobby, every activity or every interest has a publisher dedicated to it. This creates interesting possibilities. For starters you would image it’s possible to serve just on publishers (sites/apps) that have your exact target-audience as an …. audience.

Examples:

You are a horse food supplier. You want to serve ads in the about 30-70 publishers that are dedicated to horse-riding. Technically possible.

On a side note; serving ads ‘on/above’ the exact locations of actual horse stables is also a possibility.

You sell kite-surfing gear. You want to serve ads in the 25 to 45 publishers that are dedicated to kite-surfing. Technically possible. Not to mention publishers that report on wind conditions (like; Windfinder).

You are a driving instructor and want to serve ads in the 15 to 20 publishers that exist to help people learn driving. Technically possible.

You have a premium travel agency that specializes in high-end holidays in India. You want to serve ads in quality publishers that are about traveling. Also, you want to serve ads in quality news publishers (that have their own travel section. For example: cnn.com/travel) and have a ‘rich’, travel minded audience. Here is where it gets interesting.

Here is what happened:

• We started bidding at $ 21 (Dynamic) CPM – knowing that there are likely others with the same idea and there will be others bidding at/on the same inventory (a bid of $ 21 is reasonably high).

  • No traction

• We increased the bid to $ 44

  • No traction

• Increased bid to $ 111

  • No traction

• Increased bid to $ 241

  • No traction

• Increased bit to $ 900

  • No traction

You know where this is going. In the end we got traction at a bid of $ 5002 and achieved a CPM of $ 3349. Logically we applied a daily spend cap – so the client doesn’t go bankrupt. Here is something amazing we witnessed; right now, after having the campaign live for about 40 days, we have a combined CTR of 45,6% (some days this peaked to 66,67%).

What does all this tell us?

Well, first of all – the competition regarding serving in ‘quality’ publishers is constant, persistent and heavy in 2019. Secondarily; - and this something many industry leaders know – if you want to actually win millions of impressions (which is nothing in terms of buying Programmatic) – there is no way around Ron Of Network (or; RON). Essentially you could say that all substantial campaigns are relying heavily on RON. In other words; it’s possible your brand/offer will be served within ‘shit’ publishers. So what do many consider ‘shit’ publishers:

• Cleanmaster

• Flashlight-Pro

• Grindr (gay dating app)

As mentioned in a previous blogpost; these publishers are unbelievably popular when it comes down to sheer number of ad requests they fire. In other words; availability within these publishers is huge.

Rather interestingly, this client insisted on just serving within this specific whitelist. The CTR’s achieved are beyond believe. Campaign is ‘always-on’ and we will inform you of the progress (and; actual amount of inquires/conversions) of this campaign from time to time.