Chinatown

A crucial moment for premium video publishers

SpotX UHS Report FiguresSince SpotX released their findings with IHS last week there has been a lot of excitement about the adoption and growth of programmatic video, in particular in the UK. The report, “found that revenue across Europe from online video jumped from €22 million in 2012 to €375 million in 2015. IHS forecasts that €2 billion, which is over half of all online video advertising revenue, will be generated programmatically by 2020.”
These numbers are definitely exciting. However, I think they raise a secondary question and point to an important moment for suppliers of premium video inventory. Even at current levels of spend supply of premium, broadcast-quality video content is extremely constrained and that is the video inventory that marketers are after. For the first time, well, ever, as far as I can remember this is a first for online. We have always suffered from the double-edged sword of an effectively infinite pool of supply. Now, however, in this extraordinarily hot space, publishers may have some leverage.
Some ideas and observations:

Opportunity to move to a first-price auction

IPONWEB Optimal Auction PriceA lot of money has been left on the table by suppliers because of the second-price auction. Suppliers of quality video inventory can’t afford this anymore. There is a liquid market for their supply. The first publishers to INSIST on a top-bid-wins marketplace will reap significant benefits. PubMatic, SpotX, Tremor and others have the ability to deliver a lot of value to publishers. Of course there could be nuanced approaches here such as providing publishers the option to turn on and off first or second bid or creating a blended win rate, similar to what IPONWEB has suggested.

Because of fraud no one knows how much REAL inventory is out there

The blatant fraud exposed in Rob Leathern’s and then Mike Nolet’s pieces over the last fortnight have really kicked the hornet’s nest when it comes to video ad fraud, even on premium sites. According the Mike’s piece:

One fraudulent ad-unit shows so many ads that in 25-minutes I tracked:

  • 5359 HTTP calls. Yes, five thousand three hundred and fifty nine individual HTTP calls.

  • 70.19 MB of traffic. SEVENTY MEGABYTES. Thankfully I loaded this on my laptop as this one single ad unit would have used 7% of my monthly cell phone data allowance!

While the growth numbers projected by SpotX and IHS are exciting, I fear that as the ripples of these findings spread, they could be negatively impacted. Buyers do NOT want to be pumping cash into an ecosystem this dodgy. AppNexus’s recent revelation that 65% of their impressions vanished when they began filtering for fraud further underscores the scope of the issue.

Where will the growth come from

Where will this 5x growth in programmatic video spend come from? The inventory is constrained and largely sold out. Marketers are less interested in long-tail, short-format video and the jury is still very much out on the value of out-feed solutions like Teads (especially for desktop) and other ‘new video’ sources of supply.

Of course inventory is only half the equation. Price is the other. I do think there is probably quite a bit of upside to be achieved by switching to first-price auctions as mentioned above. This would at least create a real marketplace with real price signalling. Obviously though, not all this growth will come simply from better pricing the supply we’ve got.

There will be continued organic growth in the supply of premium video inventory as audiences continue to shift viewing habits but as far as we’ve seen to date this shift has lagged behind advertiser demand for those eyeballs.

The mobile issue

Users are quickly adopting mobile devices for video consumption. However, video on mobile can be a tricky issue. The screens are much smaller, user behaviour is very different (think fast scrolling), audio is often muted, there’s less screen real estate for gutters and out of page advertising, and data is frequently metered and expensive. What happens to video programmatic should we see a sudden, rapid shift to the majority of video being consumed on portable devices?

Redefining programmatic video

As far as I see it the only way that these numbers can be achieved but only with a few fundamental issues being addresses first. In addition to the fixes, what will have to happen is that “television” advertising will need to become “video” advertising. It’s like in Chinatown when Jake Gittes is told, “either you bring the water to LA or you bring LA to the water.”

There is a vast pool of supply sitting in set top boxes and televisions that is currently being classified as “television” ad spend. What will (and must) happen is that this inventory needs to be made addressable through programmatic means. The convergence  of offline tv and online video will create a huge pool of inventory for buyers to purchase and by layering on data, targeting, and programmatic delivery the efficiencies of online can be applied to the most valuable advertising supply resource in history.

In summary

  • The incredible growth numbers in programmatic video face significant hurdles
  • The supply of the valuable inventory buyers want is severely constrained and largely sold out
  • Publishers have the upper hand from a pricing perspective and need to aggressively embrace this fact. It’s a very liquid market.
  • SSPs for video inventory should move to first-price auctions to drive incremental revenue for publishers (or at least offer the option to do so)
  • It’s currently unclear where the 500% growth in the next 5 years will come from given the lack of inventory available
  • Some growth can come from increasing yields and organic growth of inventory supplies
  • The consumer shift to mobile exacerbates the issue of quality supply not being available
  • A huge opportunity exists by converging traditional tv and online ‘video’ advertising and seems to be the only viable way that this growth can be achieved.

Leave a Reply