Insurgent: How to take down Dart and Atlas

Posted by Darren on Jun 10, 2010 • (Comments)

Maybe the title of the post is hyperbolic but at least this idea could put a big dent in their agency revenues.

Note: these are my thoughts, not necessarily reflected of my employer.

We talk about agencies having a digital backbone. Yet, for the most part, the technologies are licensed. For today’s and yesterdays reasoning, this made somewhat sense. Moving forward, I think I lie on the other side of the fence and could argue that agencies or their holding co., should own/operate their infrastructure.

For todays post, let’s focus on the most essential part: the ad server and data store.

I’m skeptical of Google and MSFT, specifically with the hundreds of billions of impressions they serve collectively. It would make sense that every campaign served thru them would make them smarter. Hey Toyota, did you know that your campaign for the Toyota Camry just made Honda’s campaign for the Accord much smarter?

There is the above issue and now an important one. MSFT and GOOG can see every advertisers campaign that uses their system including cpms, impressions, conversions, etc. Being that MSFT and GOOG sell media too, they have a huge advantage if they were to use that data in the way they pitch us and price us.

Google has the ability to do the above in search extremely easily thru Google Analytics. They know every search term leading to our site and can price appropriately.

I’m not saying that MSFT and GOOG are doing any and all above but its certainly an opportunity for them.

In talking with one of the other holding company execs, he mentioned to me that we need to rethink our competition and Google is a serious threat to Madison Ave.

So I have outlined a few reasons above that we (as agencies) should be scared of our partners or frienemies. Now, let’s focus on why having this technology structure in house makes sense.

Every ad served today in display, search, rich media, online video and others puts off data exhaust. Some of this is extremely relevant for clients/campaigns and some useless. However, a good team with the data within the walls of the agency/holding co (ie like Varick Media) can reap tremendous benefits. Also, the data from within holding co or agency walls can become a friendly co-op and can become a major new business strategic advantage.

Think of tmorrows data exhaust when most media is served and tracked with a digital backbone.

While every media company out there wants to service as many advertisers as possible, different agencies have different relationships with different media companies and the opportunities brought forth vary. If agencies controlled their ad server, they could build custom integrations wit certain (not all) publishers that provide opportunities beyond where the market is today.

Ad servers will quickly morph into planning tools as well. Within the agency is where this needs to Iive. Once television ads gets served and tracked, planning will be forced into these systems (because of the sheer dollars chasing) and every agency planning team will want their own customizations. We have some tools like Quantcast that help with audience planning but imagine a custom tool of Quantcast-times-ten on the desks of all the agency staffers.

Agencies and holding companies will have to continue to differentiate and if we wanted to put a dent into Google/MSFT, then MDC, IPG, Omnicom, WPP (first mover with 24/7), Publicis, and Havas should be certainly thinking in this direction. The writing is on the wall.

The argument will be made to say that agencies cant run technology companies which i’ve publicly aligned with that in the past, but I’m going to argue it and say… Go and hire the right people. If we don’t, someone else will.

Tagged as , , , , , , , + Categorized as Advertising & Marketing, Internet & Web X.0
  • Nick Talbert
    The advertiser benefit from having multiple channels (and the corresponding data) being processed through a single interface is only now beginning to be properly realized. Advanced marketers are quickly understanding that there is a clear and apparent need to aggregate digital data into predictive solutions for advertising efforts, ideally in one robust and top-shelf tool.

    However, MSFT and Google's obvious bias and publisher revenue stream CANNOT be overlooked or dismissed. It creates an entire veil of doubt around any results or recommendations that come from their tools despite the quality or inefficiency of their respective tools

    There are other options besides DFA or Atlas. In fact there are much better choices without any speck of publisher bias. At MediaMind (formerly Eyeblaster), we have had enormous success with our cross channel serving platform. And we share the vision spelled out in this post for ad serving and technology should work within digital campaigns.

    Our platform incorporates search, display, rich, video, creative optimization and soon to be a set of planning tool and buying tools in premium and exchange-traded media. We want to put the most relevant data to the proper use in all aspects of campaign management, starting with planning all the way through re-using the ‘data exhaust’ from previous campaigns to optimize the next. Our interests are singularly aligned with advertisers and their agencies, this is very easily demonstrated in how we have designed the interface of our platform.

    By all means we encourage agencies to audit their current technology choices and look into hiring the people to build an agency controlled technology system. However, you might as well check us out in parallel. I have a feeling using MediaMind will be less expensive and more fruitful for you in the end.

    Nick Talbert
    Director, Product Marketing
    MediaMind
  • Hi Darren

    Nice comments. Here at Just Media we recently swapped out of Doubleclick and whilst remaining with an offsite solution - Mediaplex MOJO - we shifted because we did not see enough energy in development from the other platforms - especially in path to conversion metrics. That is another concern agencies should have.

    However I suspect a development will come that could change things even more dramatically. That is if client side web monitoring tools like Omniture start offering first party adserving directly linked back into client side systems. This solution actually makes much more sense for clients who then control both front end and back end data in one system (we all know that currently is a gaping hole in analytics data matching). That is a better fit for clients and agencies would then be required to become service agents to the client help accounts. There is precedents for this - we ourselves used to actively manage a Doubleclick account that one client had direct contractual control of.

    Agancy teams will still be actively involved, because most client groups will always need professional input into how the data collected can be applied to media optimizations...

    Just another thought.
  • So I think there are a lot of solutions out there to do this on your own without hiring on a technology team to physically build an adserver. For more independent direct marketers I would suggest a solution like Tagman to help with tracking and attribution modeling.

    For larger agencies a combination between Donovan/Mediabank and Appnexus or some other DSP solution as an adserver with horsepower should do the trick. You can always do research in comScore and quantcast for more demographic planning. But at the end of the day you want access to the data to make decisions and come up with your own proprietary way to evaluate and plan and Tagman and some more open sourced DSP's will give you the logs without charging you an arm and a leg.
  • exchangewire
    Darren, are you proposing that the agencies come together and build an ad server, build their own or party an agnostic player? I think personally that agencies are in danger because they are outsourcing a critical element of their business, namely ad serving, to Google/Microsoft. And now Google owns a DSP. I remember you commenting on one of my posts early last year about the inherent conflict of interest for Google being on the buy and sell side - owning an exchnage and buying platform. Well here we are. What happens now? It would seem the agencies need to reflect on the "frenemy" relationship with . How long before your clients ask the pertinent question, what exactly is your function in the media buying process and what value are you providing me. Well, when that happens I think you need to revise your whitepaper on distermediation, and include agencies as a player likely to be sidelined by the automation process and the dominance of Google.
  • Yes, Google is certainly a threat to pure media BUYING firms, not
    planning at least yet. It's Madison Aves fault as we basically funded
    their takeover ;)
  • Scott Nelson
    Insightful and ironic post, Darren. We’ve been offering technology alternatives to the velvet handcuffs of GOOG and MSFT for years and the single biggest barrier to market penetration is the entrenched approach agencies have taken with the incumbent technology providers. We all know neither Google nor Microsoft paid billions just for ad serving technology – it was the footprint that they needed to feed the data monster. As both you and Jonathan state, it is an exceptionally wise strategy. We’ve also met with every holding company in the industry and there continues to be a general lack of strategy in ad technology. The perception seems to be that providing the technology internally vs. outsourcing it will generate the same or only slightly better outcomes, so why bother? The reality is, to your point, results can increase by orders of magnitude if the buy side would exert influence and dismiss the status quo. When a 1% click rate is seen as a reference standard for display media, there is extraordinary opportunity to extract more value from the medium. (Don’t think for a moment that is lost on GOOG and MSFT.) Also, using our advertiser-centric first party technology, we’ve established empirically that third party cookie-based measurement overstates audience and understates frequencies by 50-100%. The last-click-wins and CPC economic media models cost advertisers billions every year. Advertisers and their agencies have the power because you control all of the media dollars that fund this little project we call internet advertising. To our befuddlement, we struggle to find anyone that recognizes that fact to point of acting on it.

    Scott Nelson
    Co-Founder
    TruEffect, Inc.
  • Bravo! Google has built an empire as an opaque market conduit. It is helpful to look at how they have done this in Search and because it has worked so well it is hard not to believe they will not mirror this in Display. You will get all they data you need to run and optimize. You will get easy access to scale. You will get first rate interfaces and APIs. All this is gratis. Percentage of spend for your DSP? Forget it, those days will soon be over. All Google wants is what everyone wants. The media dollars. The amazing thing you point out are not only are these free tools helping you but they serve as the backbone of google's internal optimization efforts at market making. The only thing that can stop this takeover is agencies and pubs working together. As was said yesterday at Digiday by someone "Agencies are happy to buy directly from pubs and not through exchanges." The technology to build this bridge is getting there on the demand side and we're making progress on the supply side but the elephant in the room is being overlooked and that's performance. If Google gets it better that's where the dollars will go and this is an area both Agencies and Pubs have little experience wrestling with. Rant over. :)
  • Jonathan, thanks for stoppinng bye. Working together is key.
blog comments powered by Disqus