Programmatic advertising has an efficiency problem. Tech platforms would make you believe that there is oversupply of audiences and inventory, which needs to be sifted and sorted through their bidding algorithms.   

Sure enough, the total amount of available inventory across all channels is huge. However, if you dig deeper, it becomes clear that advertising’s embedded supply constraint massively reduces actual reach and drives up costs for advertisers. Let’s break down this perception-reality gap.    

The Perception-Reality Gap Of Audience Reach   

On the one hand, walled garden inventory is so sought after that prices have been rising steadily (and steeply) for the past decade. The price effect of the supply constraint on these channels is clear.   

On the other hand, the open web seems huge. But the share of brand-safe and addressable inventory is only a small fraction of the total. This is the result of at least five factors:   

Quality Drives Screen Time And Attention  

When you add up all these factors, the number of impressions seen by real people that are correctly targeted in brand-safe environments shrinks dramatically. In other words, brands targeting via traditional methods are missing their audiences.    

The untapped advertising opportunity on the open web is huge. A report by The Trade Desk shows that over 60% of overall screen time is spent browsing the web, and while this might seem self-servingly optimistic, surveys show that the web is still the place where most people engage with content in depth. But there’s a caveat: humans do not spend any significant time on long tail sites with low-quality content, so advertisers spending their budgets and recording clicks on the web’s long tail are likely selling to bots.   

Instead, consumers spend time on the quality web. Sites with professionally created, informational or entertaining content still attract loyal communities and command more attention than they are given credit for. The perception-reality gap of audience reach creates two illusions: that these advertising opportunities are almost limitless, and that their value is limited. The reality is exactly opposite: advertising to loyal communities in quality online environments has huge value for brands and - like all premium assets - its supply is limited.       

How To Maximize Efficiency For Advertisers    

The misperception of plentiful supply and the misincentives created by the low-quality inventory have real negative impacts on audience reach, share of voice and ultimately sales. But all of the problems discussed above (non-human traffic, low-quality inventory prioritization, low match rates, fake matches and bad data) can be addressed today. And it all starts from abandoning the pretense that the open web can be treated the same way as a walled garden: where each website visitor is profiled and assigned a stable identifier.   

When every advertiser stops targeting the same small pool of users with matched IDs, efficiency grows. With more supply, audience reach expands by up to 60% driving incremental conversions. With this expanded reach, advertisers can then afford to use a stricter site inclusion list or shift to curated PMPs that prioritize quality environments, reducing exposure to fraud. The blue ocean of Safari, Firefox and iOS inventory without matched IDs opens the gate of premium environments with a low bid-density, where impressions can be purchased at lower CPMs. Advertisers can be safe in the knowledge that they are not bidding against themselves to reach the same, small pool of users. And they can replace cheap but fake reach and conversions with high quality impressions seen by real users at a competitive price point.    

Obviously, all of the above must be proved and - critics would argue - conversions without matched IDs cannot be measured. That, too, is a misconception. Conversions can be measured at cohort level without using intra-company ID matching. More importantly, campaign effectiveness can also be measured using brand lift studies and incrementality tests that correlate marketing spend directly with sales. This avoids two huge drivers of inefficiency: algorithmic optimization towards fake proxies that do not drive sales (such as clicks and lands), and optimization towards data sources that master the practice of cookie-bombing - dropping a cookie on as many impressions as possible in order to claim a conversion that was going to happen anyway.    

The programmatic advertising efficiency problem is solvable. In fact, it has already been solved and used by smart advertisers. By prioritizing advertising on quality sites, abandoning inefficient ID-matching processes and using measurement frameworks that are more directly linked to sales, advertisers are creating an efficient model that reduces waste and drives better outcomes.