How do you manage RTB quality and performance across multiple platforms and devices?

One thing i found helpful is by defining goals of our campaign to align with the business objectives. Depends on which part of the funnel these campaigns serves, marketing team can define these objectives easily (primary or shared KPI's between teams). It could be a engagement metric and a performance metric. (i wouldnt recommend going more than 2 to 3 KPI's as this will create so much noise in results). Its high time marketers should think through full funnel media strategy. By focusing on these key areas, such as: - Keep top of funnel investment addressable - Create sharable KPI between teams - Work on your media mix models - Create focus groups and gather as much as insight out of these campaigns.

It's important when you're defining KPI's to relate this back to real business goals, media KPI's should be the closest proxy to the core business KPI's. Because of this fact, often a big part of defining the most important KPI's start with talking across the silos? - How does the c-suite measure success? - How does the CFO measure success? - What do your departments goals and KPI's relate to the broader business? - Does everyone in the business agree on what the most important business KPI's are (if not we need to create some type of concensus, or decide)? The next step is to map out the events/goals that align with your business goals and start analysing to find correlations and use this as a basis are well as workshops to prioritise

Other considerations include: - Ease of data integrations ie does the DSP have adjacent analytics or audience management platforms that can feed the bidding algorithms, or facilitate custom or 3rd party bidding algorithms. - Scale and reach: It's no secret that some RTB DSP's such as Google' DV360 have access to larger reach than most, due to the scale of their SSP's and owned assets. If you want to build a sophisticated detailed account structure, then being able to guarantee reach and scale, makes this much more efficient. - How rich and accurate are the DSP's data sources? Not all data is created equally, some are going to have access to much richer, up to date data, for often free or cheaply due to the nature of their core business.

You should take an iterative approach to testing your assets. You should be getting test ideas initially from analysis of your own data? - Are there landing pages that continually out/underperform what do they have in common? Can we create hypothesis to feed our tests? Analysing your own data sources will often throw up a ton of potential hypothesis'.

One other really important point is the relationship and congruence between the landing page and creative: - Does the messaging from the creative to landing page feel natural and does it make sense? - Does the landing page compliment the creative and do they have aligned goals (an awareness campaign should not be sending someone to a buy now landing page).