Category Archives: post-impression measurement

Buying a Stairway to Attribution – Part 2 (Getting Integrated)

In the last post, Buying a Viewthrough Stairway to Attribution – Part 1, we looked at the first two steps in viewthrough measurement. These two are easy enough and come from an ad network partner or your agency’s display ad server.


Proceed to Step #3 brings together the two very different worlds of display ad servers and site-side analytics systems. At a high-level, these systems notoriously have measurement discrepancies – ad servers are typically server-side while modern site analytics solutions are all or in-part JavaScript based. In addition, there are different filters being used to exclude internal and robotic traffic.



Nonetheless, an integrated solution can be a game-changer for most digital advertising clients. Most are still struggling to understand the latent (non-click) but now quantifiable brand influence of display media on subsequent destination site behavior.

  1. Site Analytics integrated with Third-Party Ad Servers (3PAS)

    • Upside: A really powerful way to advance viewthrough measurement capabilities is to leverage the data contained in the digital media agency’s ad server. From that standpoint, the digital marketer is combining the best of both worlds: a) digital media agency-of-record’s ad server platform and b) existing in-house site analytics system-of-record. There are two distinct approaches, either taking the analytics platform and bringing it up the funnel into the ad delivery through ad tagging or make the site analytics platform talk to the ad server via integration. IBM Coremetrics takes the former approach and Omniture Genesis takes the latter. This approach is potentially very powerful because it considers all digital channels, both paid and unpaid in one place. For this reason, attribution can be considerate of all the digital channels being employed. Site analytics technologies that have provided related information to the VMC are included in the Viewthrough Capabilities Scorecard. If yours is not included get in touch.
    • Downside: Like ad networks/media vendors, ad servers can only measure what they themselves have tracked via the ads that have been delivered and their advertiser site-side tracking tags (DFA Floodlights and Atlas Universal Tags). While this avoids the single media vendor problem (now all display media being measured in one platform), its efficacy is then limited to tag proliferation on the destination site; this usually means only high-value tasks and conversions get tracked. Using a TMS (tag management platform) such as BrightTag can make this part much easier opening up deep-linked visitation and broader engagement activity. Difficulty on this approach can be high as either the integration must be configured upfront (Adobe Genesis) or there could be ongoing ad ops implications (CoreMetrics). Be careful what you wish for here as as the multiple fractional attribution methods enabled such as last-click, last-touch,now operational beget questions that have organizational/political implications. Like Steps 1 and 2, most of the technologies in Step 3 do not yet have ad-viewability solutions integrated at this time.

In Buying a Viewthrough Stairway to Attribution – Part 3, we’ll address the above point about fractional attribution. We’ll look at taking this question away altogether from human marketers and allowing the data to speak for itself through heavy datamining and algorithmic attribution modeling.

Buying a Stairway to Viewthrough – Part 1 (The Basics)

With the growing interest in defining viewthrough by multiple participants from across the ecosystem, many digital marketers are still scratching their head. This is the first of a short series of posts to provide an analytics-first view of viewthrough measurement and how you can get started.

The Stairway to Viewthrough Attribution
One of the great things about the Viewthrough Measurement Consortium (VMC) is the decentralized aspect, which enables the collection of diverse points of view. An outstanding point was made recently, which led to the thinking about the actionability of viewthrough tools and attribution approaches. That led to the conception of a viewthrough measurement progression chart, i.e. a kind of specialized Analytics Maturity Model. For digital marketers seeking to understand the complex analytics around post-impression measurement, it is more of a stairway to viewthrough attribution…Hopefully, that doesn’t offend Led Zeppelin fans too much.

Delving deeper, there are five distinct steps in the Viewthrough Stairway. What’s more, as tool capability increases so does technical complexity – it looks like this:

Here to Get Started
The first place to look is to your digital media vendors and digital media agency. The good news here is that the cost is low and the benefit potentially rich without having to do too much.

  1. Ad Networks/DSPs/Publishers
    • Upside: The simplest way to get viewthrough measurement is to leverage the display advertising and site tagging infrastructure of a media vendor, Demand Side Platform or ad network. Many of these tools have a nascent viewthrough capability but may or may not report it out explicitly. Tracking it will always require the advertiser to place tags on one or more pages on the destination site(s). Companies that have provided information to the VMC are included in the Viewthrough Capabilities Scorecard. If your ad network or media property is not included get in touch.
    • Downside: The technologies can only measure what they themselves have tracked via ad delivery and through their  tracking tags placed on the advertiser site. If the display strategy relies on just one company then this can work and still be comprehensive; it means that viewthrough is only happening across that one ad network/DSP/publisher, which is not realistic for larger enterprises. Most of these companies tend to skirt viewthrough discussion unless the client or agency brings it up. Ad viewability is not a big part of the discussion yet but promises to be. Last, this approach usually relies on 3rd party cookies, the efficacy of which for measurement can degrade over long periods of time with some prone to deliberate blocking.
  2. Third-Party Ad Server (3PAS)
    • Upside: The next best way is to leverage your digital media agency-of-record’s ad server platform. There are a few major ones to choose from: DFA/DART, Atlas, Pointroll, MediaMind and Mediaplex are the most prevalent. There are behavioral data sharing issues that arise with Google(DFA/DART) and potentially Facebook (Atlas). In this scenario the ad server reporting system is used. Like ad networks/DSPs/publisher ad sevrers, this requires one or more page tracking tags to be placed on the advertiser’s Web site. Ad servers that have provided information to the VMC are included in the Viewthrough Capabilities Scorecard. If yours is not included get in touch.
    • Downside: Like media vendors, ad servers can only measure what they themselves have tracked via ads delivered and advertiser site tracking tags, e.g. DFA Floodlights and Atlas Universal Tags. While this addresses the single media vendor problem (now all your media can be measured), its efficacy is then limited to tag proliferation and this usually means only high-value tasks and conversions get tracked. Most ad servers do not have a native ad viewability measurement capability yet, although DFA is close via the Google Ad Network. A significant issue for the raw 3PAS approach is that most viewthrough report ignores other digital channels Again, DFA has some ability to include Paid Search if DART Analytis is being used but still misses organix search, email, social, affiliates and other sources of traffic. Also, ad servers were just not built for page and user centric reporting like a site analytics tool is so de-duping is a big problem. Same issues with browser cookies.

     

In the next post, we’ll ascend to the next step in the Viewthrough Staircase. We’ll take a closer look at a more advanced approach that integrates data across agencies’ third-party ad server and advertisers’ site analytics platform for maximum convenience.