Recently I was asked the difference between traditional display placements and programmatic buying. My answer was simple. Where traditional media planning and buying was based on lists of websites and cherry picking which of those might deliver an audience based on the site user’s general profile or readership. Programmatic is buying audiences across multiple sites, so targeting individuals rather than sites. The other huge difference is the way that data sets are applied to enhance targeting opportunities and all of this is applied in a real time bidding environment across demand side platforms and ad exchanges.
Now it got a little technical towards the end so hopefully this diagram will help demonstrate the life of a programmatic ad. There are 11 steps from the moment a user enters a site to how the ad is shown to the open HTTP connection.
Hopefully, you will feel more comfortable understanding the difference between traditional media buying and the programmatic world. There are many benefits to brands and clients for investing in the programmatic way of buying media and two of the significant benefits include:
- Brands have the opportunity to reach audiences based on multiple consumer characteristics such as behaviours, location, social circles, income, age, education and occupation. Display buying has never been so targeted to specific users.
- Buying media is no longer based on a set agreed CPM or CPC across the whole media plan as programmatic buying allows the costs for each impression to be edited in real time. Depending on the optimisation this will allow a brand to target specific groups more cost efficiently than ever before.
How did Programmatic come about?
Programmatic buying is an enhancement of the real time bidding (RTB) solution combining more data sources and more DSPs and exchanges.
The Evolution of Programmatic Buying.
With so many data layers providing more sophisticated and diverse levels of targeting, programmatic planners, buyers and programmers will be swamped with data. The struggle for many agency solutions is that the reporting is an out the box solution. This by no means devalues the data available in the whizzy dashboards, but what if some clients wanted more? Do agencies currently have the talent and algorithms for custom reports? If not then the question for programmers and thought leaders in the programmatic world is: How do we find out more than what the whizzy dashboards are telling us? E.g. take exit rates on video advertising. There is a single provider I know that can provide exit rate by second rather than slots (for example: 0-24, 25-49, 49+).
Research and Test results are not widely available. For example, how does programmatic buying work with attribution modelling? Does it provide significant ROI vs. the previous methods of buying media? Please note that with all research of this nature results vary by sector. However, given the diversity in targeting options available with programmatic buying, research should also be available by audiences too. This level of transparency will substantiate the true value programmatic produces. If this level of information is not available then retailers and brands needing programmatic to produce an ROI may not wish to invest past a testing stage.