“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Benny Arbel, founder and CEO at myThings.
We’ve all seen the subject of programmatic campaign transparency hit the spotlight recently.
“When someone claims there’s not enough transparency in programmatic, I encourage you to take up the challenge and show them how far we’ve come,” Joanna O’Connell, AdExchanger’s director of research, wrote in a recent column.
I’d like to do just that.
We have come a long way with data visibility as the black box continues opening up, slowly but surely. Why is this happening? For more and more advertisers, especially the larger ones, strong performance is almost a given. They are demanding added value. The days when people in the industry didn’t care why campaigns worked – as long as they performed well – are becoming a thing of the past. As one CMO with whom I recently met phrased it, “I’ve been running retargeting for a few years now but I have learned practically nothing.”
There are two most important aspects of data visibility:
Running a brand’s ads through ad exchanges and networks means a campaign can stretch out to thousands of sites. Knowing exactly where a campaign is running on a domain level, rather than on an exchange or network level, and how each is performing, can prove highly beneficial to a brand as it pinpoints audience reaction within the context in which an ad was displayed. Ultimately, it helps marketers optimize their media buying across all channels and identify key contributors of traffic.
Another important component in media transparency is cost. Disclosing a retargeting vendor’s margins is directly correlated with the type of business model used in a campaign.
In the case of “cost plus,” the advertiser pays for the media plus a predetermined mark-up for the vendor, typically tied back to a cost-per-acquisition (CPA) goal tracked using the advertiser’s preferred measurement methodology and key performance indicators (KPI). It is therefore a transparent pricing model.
In CPA or cost-per-click (CPC) campaigns, a cost element is usually not included, as the vendor takes full risk in media buying, requiring greater flexibility to ensure sustainable profit in the long run. This has been the approach since the early days of performance advertising, which began with affiliate marketing and later evolved into personalized retargeting.
In this context, it is important to remember that in the world of performance advertising, return on investment is the No. 1 KPI. So long as the ROI target is met, media cost is either marginal or irrelevant for many advertisers.
The bottom line is that it boils down to preference. If transparency and ROI is vital for an advertiser, a cost-plus model is most suited when running a retargeting campaign. If bottom-line ROI is the focus, a CPA or CPC campaign may be more suitable.
Going forward, media transparency is expected to include more accurate viewability and ad engagement measurement as it is currently still in the early non-standardized phase, but eventually it will become a common currency. Smart machine learning will also analyze performance per media sequence, providing visibility into the optimal media consumption paths that precede a conversion.
Audience transparency informs marketers about how each of their audience segments perform within the framework of the same campaign.
Traditional programmatic audience segmentation divides users according to their interests and demographic information, or further down the funnel into cold, warm or hot segments, providing the basis for real-time decision-making. For many savvy advertisers, however, such widely defined segmentation based on one or two dimensions no longer provides sufficient visibility into their audience.
If given the option to go deeper and customize segments based on specific business goals, marketers would benefit from far deeper audience transparency. These dimensions can include a combination of parameters related to user activities and engagement, user value, geolocation and even first-party data from their customer relationship management (CRM). For example, if a marketer seeks to reactivate their high-value dormant customers, the segment can include users who did not visit the site since last Christmas but their last order value was large.
Other types of advanced audience data offered by innovative companies include performance uplift by interests, content of the page where an ad was served – as determined by a third-party provider, rather than the ad exchange – past purchases and mobile devices or operating systems, to name a few. State-of-the-art dashboards also offer customers the ability to cross dozens of parameters and independently query their own data.
By gaining access to the data of each such granular segment, a marketer can understand which creative and messages for which specific audience group generate the highest engagement and profitability. Equally important, they would also learn which do not.
Down the road, with programmatic gaining momentum and marketers becoming increasingly data-savvy, I expect greater demand for audience visibility and insights, especially via a broader use of CRM data within their campaigns, and even intertwining their databases with their most trusted vendors. It’s a key factor in understanding their omnichannel audience as it constantly moves from one device to the next, and from offline to online interactions.
Data-driven programmatic campaigns have transformed advertising. However, they should be leveraged not only to help machines make decisions but also to empower human beings who need these insights to navigate the highly complex digital marketing environment. The good news is that the black box is slowly becoming a glass box.