Does Programmatic Media Work? Prove It!

As usage grows and as more people learn about how it works, brands are starting to question the true value of programmatic media.  Lower viewability, higher ad fraud, and more remnant inventory relative to directly negotiated buys are a few sources for doubt.  All these factors have caused some brands to question if programmatic media actually helps to drive conversions — or if conversions would have been there without it.  

This is reminiscent of a common problem with paid search, the first programmatic media channel.  Branded keywords typically have the best ROI on a media plan.  But the question that always comes up is, why should brands pay to show ads on their own branded keywords if their free organic listings already show at the top of the page?  

Many search marketers respond with fears of competitors controlling their brand keywords.  But seasoned veterans run tests, pausing branded terms for a period of time to see if organic listings pick up the slack.  Most of the time, these tests show the inherent value of branded terms — though sometimes they don’t.  As with branded keywords, the true value of programmatic media can only be found by testing and proving it works.  



The first step in this process would be moving away from a last-click measurement system.  The last-click methodology typically works well for retargeting, but not for other forms of audience targeting, programmatic video, programmatic mobile, and programmatic native.  While the last click is a strong connection to a conversion, it is not the only one.  Trust needs to be built on the inherent value of seeing an ad, similar to how advertisers put trust in TV, print, and out-of-home ads.  

Viewability and ad fraud are key components of this.  The ad needs to have been seen by a live human.  Most DSPs have integrations with viewability and ad fraud vendors, and some even have them built directly into the UI for reporting and targeting.  Moving from last-click to view-through conversions is the foundation of proof.

Defining the value of an ad is the next step.  The best way to do this is to measure the lift of an ad’s performance against a placebo ad (one that does not promote the brand).  The difference between the performance of a brand’s ad and a placebo ad is the only credit a brand’s advertising should receive.  
Nielsen’s Vizu studies and Oracle’s Datalogix can measure the incrementality of a programmatic media ad by measuring the lift in offline sales.  A few DSPs have also developed placebo testing capabilities that allow advertisers to split a campaign’s exposure between a brand’s ad and a placebo.  These tools make it easier to see the incremental benefit of a programmatic ad.

Fractional contribution is the final piece that brings it all together.  Companies like VisualIQ, Adometry, and Convertro calculate the credit a programmatic media ad view should get in the path to a conversion.  This is extremely important because the touch points of most programmatic media tactics (other than retargeting) occur upstream, higher up in the conversion funnel.   

Fractional contribution redistributes credit from last-click conversion tactics, most notably search, unveiling the true value of third-party audience targeting tactics, programmatic video, programmatic mobile, and programmatic native.  DMPs have taken this a step further by integrating with most fractional contribution partners.  The integration allows advertisers to see the fractional conversion credit specific custom audiences built from DMPs should get.  This empowers advertisers to have more confidence in the value of the audiences they are targeting.

This measurement structure requires funding that won’t necessary go directly towards driving a brand’s KPIs.  It also takes time to bear results.  But it’s a necessary investment to understand how much advertisers should actually be spending on programmatic media.

2 comments about "Does Programmatic Media Work? Prove It!".
Check to receive email when comments are posted.
  1. Jon Schwartz from Netmining, September 11, 2015 at 1:54 p.m.

    If you are only looking at programmatic display as a direct response ad strategy, then you're really not looking at it the right way or you're certainly looking at it too narrowly.  Show me how a specific TV ad in prime time can be directly attributable to an increase in sales or a specific sale to a specific customer or even household?  Does that stop marketers from spending $80B per year on TV?  Nope.

  2. Chris Quinn from TripleLift, September 12, 2015 at 12:16 p.m.

    Brian - very well written and I agree with your general sentiments.  I would, however, urge you to place the same scrutiny on the (IMO largely imperfect) measurement providers as you do on the buying platforms.  To crown measurement partners like OBE and DLX as absolute truth while scrutinizing the buying partners is quite one-sided. 

Next story loading loading..