Opinion: Year of the Snake puts the squeeze on ad tech

Recent data suggests that the aggregate success rate of online advertising is around 0.1 per cent. With the amount of data, contextual content targeting and real-time technology now available to us, I would suggest this amounts to a massive failure in our industry. Hundreds of millions of dollars are being invested in the industry across real-time bidding platforms, DMPs, DSPs, and data exchanges, all with the promise of delivering greater scale, efficiency and results for brands.

As an ad tech insider, its hard to keep up with the ever-changing landscape, so I imagine this is significantly harder for advertisers themselves. Last year, real-time bidding (RTB) ad spending grew 98 per cent from 2011 to almost US$2 billion dollars in the US and is expected to continue this growth trajectory through 2013. Our own studies at Experian show that it's also growing in Asia, although growth is challenged without the rich data supplies and cookie pools of Western markets.

For advertisers, RTB promises to create efficiency and increase targeting capabilities, allowing them to reach their intended audiences and optimise their budgets at the same time. As inventory supply increases, CPM tends to decrease. Whilst this is good news for advertisers, it doesn't work for premium publishers, which is why many still shy away from RTB. Publishers who guarantee inventory on their premium sites can therefore still command a higher CPM as they control the price directly. And whilst guaranteed inventory may be important for advertisers, it means they lose the advantage of real-time targeting and therefore capturing consumers that are 'in market'. Guaranteed inventory sales is a largely manual and multi-step process that largely doesn't utilise the programmatic methods of RTB, and therefore nowhere near real time.

Today, RTB platforms are largely used for remnant inventory, as they can efficiently arbitrage the price based on supply and demand, but premium publishers won't sell their top inventory on an ad exchange when they can't control price. According to a Casale Media survey published late last year, media buyers want to see greater emphasis on inventory quality, more guarantees for brand safety, and greater transparency in reporting. Most publishers (84 per cent) are concerned RTB will devalue or commoditise their inventory.

One of the biggest challenges with RTB today is that its ROI is based on targeting audiences based on fresh intent, in exactly the same way search does. But the reality is that it isn't anywhere near as real-time as search. RTB ads are targeted using cookies, and the behavioural data they hold may in fact be days or weeks old—well beyond the moment of truth at which time the consumers are in-market, ready to buy.

This isn't to say that non-real-time ads are useless. A study published by the University of Maryland last year demonstrated that past exposure to online ad campaigns does significantly impact user response, in addition to their intent. Using real ad campaign data, the researchers were able to create a model using both parameters that better identifies target users compared to conventional targeting models. Told you so, I hear some traditional marketers say.

Another point of difference is that nearly half of all search results pages don't display ads, as the search engine determines in real-time that the query has no commercial intent. And yet in the display market, we serve up ads on every page, regardless of relevance, because the technology lets us do so. I suggest this is a significant reason behind the 0.1 per cent response rate.

Today, RTB constitutes a small but growing component of the online display market. I'm confident that many of the issues discussed above will be addressed in time by a few of the many vendors in the ad tech space who will emerge as winners. The bigger question for me is how this translates into the mobile ad world. The lack of a universal cookie alternative on mobiles creates a significant, and as yet unresolved, technological hurdle as to how RTB will work on mobile handsets. Whilst there are options being developed, none have reached the tipping point of becoming a standard that enables exchanges and DSPs to reliably identify mobile users, other than using location as the only context.

And of course, the jury is still largely undecided whether mobile banner ads actually work as an effective advertising medium, which is quite the conundrum given the migration of web traffic to mobiles. I predict the Year of the Snake will put a python-like squeeze on the ad tech landscape as many vendors struggle to meet the high expectations of their investors. The choice for advertisers next year will be simpler.

John Merakovsky can be followed on Twitter @Experian.