Transparency In Programmatic And Lessons From Ad/Fin’s Demise

"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 Christian Polman, chief strategy officer at Ebiquity.

The bold move made by ad tech business Ad/Fin to drive transparency in programmatic media trading came to an abrupt end last month with the closure of its business. Ad/Fin’s promise to centralize programmatic data across the supply chain was admirable. More recently, Ad/Fin shifted to a reseller model to supply its solution to several brands by way of consultancies. But it was seemingly too little, too late.

Two years ago, we partnered with Ad/Fin, the Association of National Advertisers and the Association of Canadian Advertisers to produce an in-depth study of transparency in programmatic. The study analyzed hundreds of campaigns and billions of impressions. From this research, we outlined an 11-step playbook for advertisers to enhance accountability and transparency in programmatic media trading, which remain highly relevant today.

The study also produced broader lessons about achieving transparency in programmatic.

First, we learned how challenging it is to trace transactions across the supply chain, let alone down to the granular log level of individual transactions, as Ad/Fin was undertaking. This is particularly true in the context of nondisclosed programmatic buying agreements.

Second, we learned the logistical, technical and financial challenges associated with storing vast amounts of log-level data. Programmatic ad impressions can generate more than a hundred data points each, so a comparatively modest campaign with 10 million impressions could generate as many as 1 billion individual data points. This is clearly costly to run and not necessarily required to identify key issues and performance opportunities. Many of the same insights generated by log-level analysis can more easily be extracted using readily available APIs or standard data exports.

Since the study, we’ve also demonstrated that advertisers need to take a holistic approach to solving the common issues in programmatic. The challenge with a log-level approach is it starts with the data rather than with insights or hypotheses. Instead of starting with granular data, it’s much better for advertisers to first focus on business outcomes, such as return on investment. This is even more powerful when complemented by analysis and benchmarking of buying KPIs that will enhance performance. And from there, brands can delve deeper into their programmatic tech stack to uncover further inefficiencies.

As with any complex problem, the solution requires a broad approach, and inevitably stretches across legal, process and technical disciplines. Furthermore, progress is limited if part of the solution doesn’t involve aligning partners to the desired outcomes and KPIs.

Through hundreds of digital assignments and by analyzing more than 3 billion euros in digital spend, we have learned that while disclosure and transparency can help advertisers to achieve greater control, in and of themselves they don’t guarantee that programmatic media buying will deliver better results. Alternatively, monitoring, reviewing and benchmarking performance focuses instead on opportunities to drive better value and outcomes.

Improving the performance of digital spend is critical in today’s complex and fragmented ecosystem. Ultimately – as the market has shown with Ad/Fin’s collapse – it takes a comprehensive approach, focused on insights and analysis, that’s closely connected to business objectives.

Unfortunately for Ad/Fin, a sole focus on granular data meant it could never quite solve brand challenges in programmatic media alone, and ultimately, it lost sight of the bigger picture.

Follow Ebiquity (@ebiquityglobal) and AdExchanger (@adexchanger) on Twitter.

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>