• http://www.facebook.com/maryannasophia MarYanna Sophia

    Hi, first of all great article. second, could you please rewrite for me this sentence: direct brand and agency contacts buying display media …, is a little confusing for me and I will like to analyze the article complete and understand it fully. Thanks.

  • http://www.facebook.com/jrstew2 John Stewart

    Great read. I appreciate your use of humor and the infographic to illustrate your point.

  • http://twitter.com/vantagelocal Vantage Local

    Thank you for making this point so well.

  • http://www.facebook.com/marc.poirier Marc Poirier

    Hi James. I think the problem goes both ways. Taking 100% of the credit for a conversion based on an ad impression is even more ridiculous than taking 100% of the credit for a click. Giving ad impressions the proper weight is key in solving this conundrum.

  • http://makethemclick.com.au/library Mark @ Make Them Click

    The problem with all this “industry” opinion is that it conveniently forgets what the client wants. The client wants a result. They want to see money flowing into their bank form an ad campaign. If they can’t see a direct correlation, they won’t keep advertising regardless of what new buzzword industry types come up with.

    It’s a pretty simple concept that seems to elude most agencies.

  • http://twitter.com/KennethKarl Kenneth Karl Nielsen

    My opinion as a media buyer is that salesmen often try to “oversell” the value og view through conversion. Ex. taking credit for ALL sales by people, who have seen the creatives. I stil work with finding the correct/best attribution for bannerview, and it seems to me, that this is probably the solution,

  • http://twitter.com/celwell Chris Elwell


  • GS1981

    Agree wholeheartedly with the main thrust of this article, but I’m not sure the solution is conversion attribution modelling. In fact I’d go as far as saying those tools are simply another ’emperor’s clothes’ solution designed to keep marketing money with the companies sitting at the end of the funnel, rather than being spent on the most effective solutions for driving purchase decisions.
    Generally speaking, attribution tools assign a higher contribution percentage the closer an advertising event – click or impression – is to the conversion event (i.e. most credit goes to the event in last place). But this creates yet another disproportionate perception of performance – surely credit should be given for events appearing in the right place, rather than in the last place? If someone has already decided to buy a product due to various influences, they would often purchase via search – so why should the final search event get more conversion credit than the media owner who targeted the right user with the right message at the right time, having much greater influence on the conversion? Why should display vendors who simply place banners behind the search links people visit after they have already decided to buy, get more credit than the display media who influenced the prior decision to buy? End of funnel solutions do steer a conversion to a particular vendor (do I buy the DVD from shop A or B), thus offering clear value when fighting competitors for ‘fallen apples’, but they do not create the conversion in the first place (I want to buy the DVD), yet under attribution modelling or viewthrough / clickthrough models the last-event display & search firms are always paid or credited more.

    It may not appeal to marketers to accept that online display media is not used by audiences as a search, navigation or buying tool (despite it overwhelmingly being planned, bought and measured as if it were), as it is simply the same ‘subliminal’ beast as TV – nothing more and nothing less. It doesn’t have the tangible benefit of print, where users save or revisit ads they saw earlier – like TV, once user consumption progresses away from banner exposure, they can’t go back to it. They consume the ad in passing. It may not be video, but it’s still only visible for a matter of seconds. And guess what – if someone does remember an ad and subsequently investigates, search gets the credit!

    Arguably, for sales campaigns we should at most just be measuring the number of contributions made (not conversions, contributions). Not weighted conversions, not last event conversions, not path position; just the number of conversions a media delivered, weighted against the cost or volume of inventory used by that media (within a relevant conversion window). Then if you’re using a CPA / CPC display media, measure their performance against search (the media they’re competing with) rather than other display media (who they’re set up to take money from, by claiming to do the same job more effectively, when really they’re doing a different job wiht the same tools).

    It is really very easy to measure display media in a manner which reflects the job it is meant to do, rather than measuring it’s effectiveness at something it should not be expected to do. Display media is not and has never been broken – the process of attributing value to display media is where the problem lies.