I recently met with the planning and media teams of one of the largest CPG brands in one of the largest holding group agencies in our industry. During this meeting we discussed ways mobile can be leveraged to better understand audiences, as well as driving and measuring sales. The entire time, I was wondering when the viewability question would pop up. Just when we started summarizing the action items, it came from one of the media planners.
“We want to make sure you can deliver 100% guaranteed mobile viewability.”
As all media sellers reading this know, this is a loaded question. If you say you cannot support 100% viewability, you are off the plan. If you start explaining that it’s complex, as measuring viewability for mobile is in very early days, you are off the plan. And if you say you can support it, you are setting the wrong expectation that will yield a quick win, but likely foster long-term disappointment.
So let’s start with the fact – you cannot guarantee 100% viewability at scale in mobile programmatic advertising in 2016. The truth is that you can only measure significant chunks of inventory using an MRAID-based solution, and very small inventory pockets using an SDK-based solution.
Herein lies the secret that nobody talks about – it is not just that you cannot buy 100% viewable inventory, you also cannot currently measure the majority of mobile inventory. It is, in fact, impossible. By way of example, a mobile impression is considered to be measured only if the user has viewed it for more than a second. An Android user can press the back button to close the ad after 0.5 seconds. In this case the only option for the entire industry is not to count it at all, and also not to pay the publisher.
There are two methods to measure viewability for mobile, either via SDK- or MRAID-based tracking. SDK-based tracking is an accurate technology that can identify exactly what is in view and what is not in view. The other big advantage is that it can also measure VAST and native video.
The problem with SDK-based tracking is that the process of implementing an SDK is slow. It can take exchanges up to 6 months to implement. By the time they are distributed, they will need to start a new cycle of implementations. More than that, not all exchanges will use the same vendors, so you will never have consistency with measuring viewability at scale. As of May 2016, none of the major open exchanges have implemented an SDK-based tracking solution for their SDK’s.
MRAID-based tracking, on the other hand, requires no major implementation, and it can work across exchanges. However, it is only supported on MRAID-based impressions. Once MRAID3 is adopted, this will become a more scalable and accurate way of measuring viewability.
So, what do we do in the meantime?
Let’s start setting the right expectations and be more transparent about what we can offer now. Viewability vendors, exchanges and DSP’s should be very loud and clear that we are not ready yet to deliver based on 100% viewability for mobile in-app. Viewability doesn’t need to be approached like it’s an insurmountable problem. This is especially true in mobile, where we have the benefit of in-app inventory that usually features a single ad space per screen, and inventory is also usually above the fold.
In my opinion, viewability, while a hot topic, is not the most important topic facing our industry today. Mobile presents amazing opportunities, but the reality is that this industry is still new and young. Concepts such as non-human mobile traffic and location validation are two of the more pivotal topics just scratching the surface of the opportunities and challenges that lie ahead. However, unless we establish a valid demarcation point right now around viewability measurement limitations on mobile, we will lose our ability to demonstrate progress.