Real Data Shows Real Slowdown In Online Display, More Expansion In Total Media Buys

In what is likely the most accurate accounting ever of the U.S. media-buying marketplace, a new, more realistic and somewhat surprising picture of ad spending is beginning to emerge. The data -- the first ever to be released publicly from a system that taps directly into the data processing systems of Madison Avenue’s major agency holding companies -- is shedding light on real market behavior, including a dramatic slowdown in the expansion of online’s premium display advertising marketplace, and a corresponding upsurge in so-called “secondary” display media-buying channels such as ad networks and exchanges. 
The data, which comes from Standard Media Index (SMI), startup that still may not be familiar to some big agencies, already is processing -- and aggregating -- actual transaction data from the Aegis, Havas, Interpublic and Publicis media organizations, competes in a market that previously relied on estimates derived by applying average media cost estimates to media buys monitored by third-party media tracking firms such as WPP’s Kantar unit and Nielsen (see related story in today’s edition of MediaDailyNews).
“We saw this trend with [online] display flattening out three or four months ago,” says Sue Fennessy, global CEO of SMI, who co-founded the company five years ago in Australia, which is where she says the trend first began to manifest in SMI’s database. Because SMI already is tracking nearly 100% of the media-buying data from agencies Down Under, she says the data is extremely accurate and a solid bellwether for trends in other markets.
Based on its first year-over-year release of U.S. media-buying data, SMI finds that the U.S. online display advertising marketplace grew only 4% in 2012, a rate that is more than 40% slower than the 6.9% rate SMI found the total U.S. media-buying marketplace expanded for all media in 2012. While that total rate is much higher than estimates released from other sources, including the major agency holding companies, SMI’s is the first ever to be based on actual transaction data pooled from a critical mass of big agencies. That said, it should be noted that even all of the major agency holding companies combined do not represent all, or even necessarily the majority of ad spending for many media -- especially digital, where so-called “long-tail” advertisers represent a major share.
Nonetheless, SMI’s data is a revelation, because it is the most empirical snapshot of actual U.S. media-buying behavior among the major agencies ever released. Among other things, it reveals that secondary display advertising markets, including programmatic exchanges and third-party ad networks, appear to be slowing premium display’s growth as advertisers and agencies shift in favor of more efficient inventory.
While the share of such secondary markets is still small -- only a few percentage points of total digital media buys placed by big agencies -- the efficiencies are much greater both in terms of labor and inventory costs. Not surprisingly, the volume of display advertising inventory purchased through exchanges grew nearly 50%, while demand from ad networks -- a relatively more mature source of display ad inventory -- expanded nearly 14% in 2012.
When other rapidly growing digital sectors such as search (+19.6%), social (+31.0%), mobile (+20.6%), and email (+54.9%) are factored in, total digital media buys grew 15%, more than double the overall expansion in U.S. media buys during 2012.



8 comments about "Real Data Shows Real Slowdown In Online Display, More Expansion In Total Media Buys".
Check to receive email when comments are posted.
  1. Henry Blaufox from Dragon360, March 6, 2013 at 9:25 a.m.

    The SMI data document the extent of growth in "more efficient" ad inventory sources, such as the exchanges. But how much of those buys are guaranteed to be seen, and how accurate is conversion / funnel attribution? when many sources can still claim some credit for the same ad and some resulting conversions?

    What is the quality fo that inventory and the auience being delivered?

  2. rachel flow from mediaocean, March 6, 2013 at 9:46 a.m.

    Mediaoceans Programmatic MBuy platform targets the audience with access to 85% of all available inventory merged with 3rd party data to ensure effectiveness. Attribution is identified by tags so it is accurate. Reporting shows CTR on impressions served along with conversion data. MBuy direct can also help you place direct buys ensuring specific site targeting and recommendations for sites due to your targeting preferences.. all on one platform eliminating your need to rfp all the sites you want

  3. Joe Mandese from MediaPost, March 6, 2013 at 10:13 a.m.

    @Henry: Great questions. I don't think the SMI data seeks to answer them. It just explains what the overall market behavior -- and share shifts are. What you are getting it is probably better answered by other people -- maybe even some intrepid trade journalists.

  4. Joshua Chasin from VideoAmp, March 6, 2013 at 10:57 a.m.

    Hey Joe. The data here is described as "volume of inventory" and "share of buys." are these data based on dollars, impressions, or something else?

  5. Joe Mandese from MediaPost, March 6, 2013 at 11:31 a.m.

    @Josh: It is actual transaction data of the media buys processed by agencies with the media. The shares are based on the inventory costs (ad spending) not on the impressions delivered. Seems like a complement to comScore or others.

  6. John Grono from GAP Research, March 6, 2013 at 5:30 p.m.

    Josh, it is based on 'financial' bookings. That is, it is the dollar value invoiced to the client by media owner as booked through a media agency (well those that are in the SMI pool - which in Australia is 100% of the big 'uns). It doesn't include direct sales from media owners. There can be some minor vagaries, especially in the digital world where 'bundled' sales are more common, and the allocation across channel (SEO, display, video etc. can be more guesstimate than actual - but the total is always correct). Also, timing of the loading of the data can sometimes through things out month-to-month (e.g. invoicing could be late) but across a year it is pretty darned solid.

  7. John Grono from GAP Research, March 6, 2013 at 5:31 p.m.

    Oops .. that should be "throw things out". Me bad.

  8. Joe Mandese from MediaPost, March 6, 2013 at 5:37 p.m.

    @John, Thanks for explaining it better than me (and for providing more details).

Next story loading loading..