Interactive Ads – Helping Solve the Mobile Attribution Problem

opinionThe following is a guest contributed post to MMW from Jeff Marshall, co-founder and CEO of CrossInstall.

The mobile gaming industry is often considered to be at the forefront of performance marketing. Gaming marketers have developed sophisticated methods for tracking the lifetime value of every paid user they’ve acquired. As emerging ad formats such as playable ads become more common, it becomes necessary to reconsider how attribution methods can evolve as the industry does.

There are two predominate attribution methods employed on mobile ad campaigns: click-through attribution and view-through attribution.

  • Click-through attribution will attribute a conversion event to the vendor that last led to a user clicking the advertiser’s campaign on their device.
  • View-through attribution will attribute a conversion if a user views the campaign’s creative even if the user doesn’t click through to the landing page and take the conversion action right away.

One could argue that advertisers should turn to view-through attribution methods to try to get a clearer picture of the customer journey and the efficacy of video or display ads in influencing the final conversion.  However, mobile advertisers tend to prefer click-through attribution as it feels the most intuitively tied to measuring a concrete action taken by a user to convert on a campaign.  In practice, advertisers are using a blend of attribution methods across different vendors because many mobile video networks force advertisers to use view-through attribution, while advertisers default to click-through attribution for the rest of their vendors.

One problem with view-through attribution is that it is inexact and relies on the accuracy of key assumptions.  View-through attribution fires an attribution impression even if a user did not pay attention to an ad, like a video. Also, using multiple vendors configured with different attribution methods can result in attribution-stealing. If an impression is served (but not clicked) during a campaign, view-through attribution can credit that impression even if a different vendor registered an actual click that led to a download earlier (install campaigns do not attribute the conversion until a user finally opens the mobile application, which may happen some days later).   Running a campaign with both view-through and click-through attribution can lead to misinformation and obscure where an advertiser is truly spending the most effective dollars.  Facebook in particular is problematic for advertisers working with multiple vendors because Facebook uses both click-based (for 28 days) and view-through based (for 1 day) attribution.

One method of mobile attribution fraud is to place an attribution click URL as a pixel on every ad impression served.  This allows a fraudster to register a huge number of fake clicks to pick up attribution from any users that go on to organically convert later, or to steal attribution from vendors that already legitimately influenced the user’s actions.  Ironically, this is what view-through attribution is essentially doing, albeit with a shorter lookback window.

So although some argue view-through attribution may paint a more accurate picture of the influence ad impressions are having on a user’s actions, it has drawbacks that have made it unappealing for advertisers to use widely on mobile, except where they are forced to use it by video networks or Facebook.

But what if there was an ad format that could enable a hybrid type of attribution–generating a definitive click-like event when the user just engages with the ad?  Such an ad format is now emerging on mobile: playable or interactive ads. Interactive ads are ad units that enable the user to play a game or preview an app’s experience without having to download it.  They are the mobile app equivalent of taking a car for a test drive.

As a result, developers have found that playable ads have higher conversion rates than other types of ad units.  Equally as important, installs from playable ads have proven to have higher retention rates, ARPI, and user LTVs.  Such post-install KPIs suggest that users who engage with playable ads are likely to become higher-value users – even if they don’t install the game right away.

So one way all types of advertisers can take advantage of an interactive ad format is to be able to fire an attribution (click-like) event after a user has clearly interacted with the ad all the way to the end of the ad experience – for example, completing a level.

The attribution and measurement problem in mobile is complex, and there is no single solution.  Right now the ability to target and engage customers in new ways outpaces the ability to measure the effectiveness of these techniques.  Last click attribution methods don’t help advertisers model the influence of ad impressions, and they are susceptible to fraud by bots firing fake click events.  View-through attribution is arguably even more flawed: final conversions can be unfairly credited to any impression that occurs during the lookback window, even if an earlier ad click actually led to the install.   In either case, the advertiser ends up wasting money and at the same time comes back with false conclusions about the efficacy of their campaign.  An interaction-based attribution method has the potential to deliver what view-through methods unsuccessfully strive to do: enable a real, quantifiable measure of how a customer engages with an ad.