RTB: Applying Art to Science

23 Oct, 2013

By Tony Evans, Corporate Development Director at Crimtan.

Apparently ‘math men’ have replaced ‘Mad men’. So many articles have suggested that online display advertising is now all about algorithms, RTB and programmatic buying that marketers think online advertising is something that only techies in agencies can manage – and that good results are dependent on impenetrable ‘black box’ solutions that drive their campaigns.

It is certainly true that recent technology developments have significantly improved the effectiveness of online campaigns – and that maths and science play a large part in that success, but marketers still control the budget so it’s vital that they continue to be involved in analysis and decision making. Crucially, marketers need to understand why certain strategies work so they can use this knowledge to ensure their campaigns work next time too.

Thankfully, we are beginning to emerge from a dark era where mysterious automated technologies and unfamiliar acronyms have marginalised marketers, into a more open and transparent period where the art of marketing can once again be recognised as an essential element in the success of a campaign.

Here are a few myths about RTB and programmatic buying that show why marketers should play an active part in the decision making process for online advertising campaigns.

Effective RTB is all about machine learning – humans are redundant.

For any real-time bid there are thousands – even millions – of potential variables that affect the decision to buy an impression or not and it’s true that only machines can handle that amount of data, so they are essential in managing this quantitative element. But it’s important not to completely rely on the machine because different tactics need to be used for different objectives like retargeting and prospecting and personal experience is essential to make sure other factors like seasonality, sales, competitive activity and offline activity are taken into account. For example, the machine may see that a campaign works best in a certain area, but the reason for this may be due to local offline media being used as part of the campaign. Or perhaps a current TV campaign is driving more people online. The machine may think that running the same creative/time/geo will work next time, but only a human can take the bigger picture into account.

RTB is only a direct response tool.

When RTB was first introduced all inventory in exchanges was remnant and, although it could be bought cheaply, there was a big question mark over quality. Understandably, brands didn’t want to be seen on dodgy sites and kept their distance. Now, because visitor data can add value, rates have improved, private exchanges deliver extra security and premium sites have started to make their inventory available for RTB. On top of this, new tools like brand safety protection, viewability and inventory quality scoring mean that RTB is increasingly suitable for brands. That puts marketers right back in the driving seat.

RTB is too complicated for planning and analysis.

There are some companies who offer ‘black box’ solutions and just provide the result of the campaign. For marketers to be able to plan their marketing strategy effectively it’s essential that they understand what worked, what didn’t and how the campaign was managed to produce the result. To allocate budget effectively you need to compare tactics and results. Was audience data more effective than geo and did demographic data make a difference? What inventory and contextual environments worked best and was retargeting cost-effective? Marketers need to choose suppliers based on their operational transparency and ability to answer these questions – and more. All this data is available, and marketers need to obtain and analyse it to understand where their budget works hardest.

Best practice in RTB means having talented marketers making decisions that influence RTB strategies and other campaign tactics. You need to know why the campaign worked and understand how you can use this information, so beware of companies that say their tech is so clever it doesn’t need any human intervention. The art of marketing is as important in online as in any other part of the media plan, so don’t let anyone tell you that all you have to do is hand over your money and the science will do the rest.

This article appeared in Marketing.