RTB still isn’t the method of choice for many types of campaigns. Brand campaigns in particular are still rarely booked on automated platforms; it’s interesting therefore that there is so much excitement around the concept of ‘premium programmatic’.

It’s not hard to see why. Programmatic is growing fast, while brand is where the real money lies. Anyone who can combine the two is golden.

On paper, there are many potential benefits brand campaigns can gain from automation other than simply reducing paperwork; primary amongst these is probably the improved access to the targeting options and audience metrics that these platforms offer.  

There are complications too – a manual, personal approach allows for off price-list bespoke products to be developed specifically to meet a brief.  While private exchanges mitigate for this to some extent, it’s clear there will always be aspects of brand campaigns that require manual input.

Putting these complications aside, there’s a more fundamental flaw with the underlying platforms that makes true automation currently impossible for brand campaigns.  Regardless of how many ‘premium’ solutions are being touted, if they can’t fix this, they aren’t for brand.

The problem is that all these platforms optimise yield.  Yield optimisation needs a trigger to determine the success of any displayed impression, and that trigger is currently limited to clicks. Clicks are the right goal for performance campaigns, but for brand campaigns, they are meaningless.

The goal of brand campaigns is to change consumers sentiment. Consumer sentiment exists only within a consumers mind.  No automated platform, however, is yet able to read consumers minds.  

The only trustworthy way to know what a consumer is thinking is to ask them. For this, advertisers typically use post campaign research, but, as you may be aware, post campaign research tends to happen post campaign…

This, of course, makes it somewhat tricky for any platform to optimise against while the campaign is still running.

Premium Programmatic is a noble goal, but it’s a tough challenge, and innovation needs to happen before it can be considered solved.  The theory is simple enough; the bidding platforms need to use triggers other than clicks to optimise.  

In practice, lots of proxies for consumer sentiment have been offered as solutions, but none have been reliable. A solution is required that determines accurately consumer sentiment towards a brand, and can optimise using this as a trigger.  

Does such a solution exist? Not yet.