January 15, 2013
Editor’s Notes: This article reflects a speech given by Mediasmith Founder and CEO, David L. Smith, at I-COM in Rome and was first published in iMediaConnection on December 17, 2012.
Ten years ago, Ad servers represented the total or at least the vast majority of agency tech budgets. Today, this component is less than 10% of the total media tech spend for most digital agencies. This is the LUMAscape we are all familiar with. It shows the complexity of the digital display marketplace.
Ad serving today is a small part of this overall pie.
Innovation came slow to this category while pricing dropped. Standalone companies began to crop up that should have been ad serving features.
There are a number of companies that have cropped up that arguably could have been introduced as features of the third party ad server (3PAS) had they innovated. Most notable is the DSP. For many agencies, the DSP has 10x or more of the spend of the ad server. Some of the categories of companies that could have been 3PAS features include:
Rich media: major companies like mediamind exist only because 3PAS companies did not sufficiently develop rich media. This includes not only the creative vendors, but survey vendors, integration of survey data into the cookie stream and creative content management systems
Attribution: connections between disparate sources for web display, search, social, site side is still an expensive effort affordable to only the largest advertisers from companies like Marketshare, VisualIQ and MMA.
Tag Management: Tagging systems were virtually invented for 3PAS, yet we need companies like BrightTag to bring this capability to fruition
Optimization: It took the advent of DSPs like Mediamath to give us advertiser side optimization beyond the click. We still cannot share view through with sites for optimization on regular buys. And we still need creative auto-optimizers. Many media directors will admit that they find optimization by people still superior to automation.
Verification: This whole category exists only to establish if the ad was served correctly. 3PAS should have done this. It is a feature, not a business. Companies like DoubleVerify, Adsafe, Adometry, Media Trust and now comScore do this for us.
DSP: The 3PAS company was, for the most part, our exclusive technology vendor. They could have brought real time bidding interfaces, bid management systems for SEM, etc. to us. Instead, the DSP (along with media management companies) have the largest part of our tech budgets. Again something that could have been a feature. MediaMath, Invite Media, DataXu and Triggit are all good examples here.
Targeting and Data Aggregators: targeting information is collected every day from companies using social or lookalike targeting tags inserted within the 3PAS envelope. Companies like Quantcast, Media6degrees and 33Across are good representatives of the new targeting capability. Companies like BlueKai and Exelate are leaders in this category
Retargeting: Invented by Doubleclick with Boomerang. Perfected by other companies due to Doubleclick’s failure. Now everyone does it.
R/F and GRP metrics (Measurement and Analytics): The third party ad servers were at the table but failed to follow the lead of the ARF in the early part of the last decade, (the advice to fuse data with panel companies) . This has left GRP development to traditional research companies like Nielsen and comScore. They have taken over the GRP calculation from the 3PAS, the only source of census data on a campaign.
RFP/workflow (Media Management Systems): still a major pain point for many agencies. Sites fill out agency RFP spreadsheets and too much insertion into the 3PAS trafficking is done manually. Systems started by 3PAS to do both this and publish site side specs and standards were never kept up to date. Mediaocean is a solution only available affordably to the largest agencies. Although their mBuy division shows promise for smaller agencies.
Content: No reason why site side CMS cannot be powered by and ad server like technology to personalize site experiences. Companies like Zite are doing this today. In addition, companies like Peer39 and proximic assist in contextual targeting.
Here is the same list graphically. A large percentage of the overall LUMAscape is covered by this list.
Here is the look of the LUMAscape if there were consolidation under the 3PAS. While this is a simplistic look at this, you hopefully get the point.
What’s next? Will the DSP or the media management technologies take over as the primary agency tech partner? Will Doubleclick reemerge as a factor due to the acquisitions and roll-ups being created by their now owner, Google? They have at least maintained strong market share of publishers despite the fact that Google bought them. And Atlas may be sold for virtually scrap. Still, a great deal of the ad server function today seems to be trafficking, cookie distribution and management/tag distribution. Will the need for ad serving disappear as newer technologies include ad serving as a feature rather than the other way around? What will a roll up of services look like and who will drive it? Should be an interesting time seeing what happens next.
Mediasmith also would also like to thank Terry Kawaja of LUMA Parters for permision to use the LUMAscape in the manner shown above.