It’s a well-known fact that people spend the vast majority of their mobile phone time in-app. Despite this clear opportunity, spending on in-app video advertising lags far behind web-based video advertising. Why have advertisers and demand-side platforms (DSPs) largely avoided the potential of this huge captive audience for so long?
We’ve mentioned elsewhere on this blog some of the important differences between web-based and in-app advertising, and today we will cover another – render rate. Render rate tells us the percentage of ads that are actually shown to users, and it is calculated as viewed impressions divided by ad calls filled (or programmatic auctions won).
Thanks to mature technologies, longer session times, and different user behaviors, web-based video render rates far outpace those of in-app, especially in-app outstream video.
Render rates are disproportionately low for in-app outstream video. Why?
Two common ad call practices cause in-app outstream video render rates to plummet if not managed correctly. The first is aggressive pre-fetching of ad creative. Because app session times are short and user actions are often quite brisk (think scrolling a newsfeed or swiping between videos), most video ads are prepared before the user even opens the app. Downloading the creative in advance makes the video available to start playing immediately, increasing the chances of catching the user’s attention and earning a quality impression. The problem arises when the SSP prefetches ads that don’t become impressions. Each filled ad call costs the DSP or ad network a transaction fee, so a fill that doesn’t become an impression is not only a missed opportunity, it’s also costly. If this practice isn’t improved, the demand partners may choose to discontinue the working relationship with that publisher.
The second faulty practice is misuse of parallel ad calls. Originally developed to increase fill rates, parallel ad calls place simultaneous calls to all available demand partners for each placement with only the highest-priced or best-matching bid winning the placement. Similar to the issue with pre-fetching, this can result in numerous filled ad calls that don’t become impressions. Once again, demand partners are spending money to pursue placements with very low render rates, which puts strain on the working relationship with the publisher and SSP.
Render rates can be balanced with the right technologies in place.
At Intowow, we have developed automated solutions to counter these two challenges. Pre-fetching can be made much more efficient through the use of device-level data to anticipate user patterns that determine which placements are most likely to offer impression opportunities. Our decentralized SDK technology learns user behavior from each device to pre-fetch ads only for the placements that are most likely to generate impressions. This approach maintains value for the advertiser and supports a healthy relationship with the publisher, as well.
To better manage parallel ad calls without sacrificing fill rates, we have developed a smart firing scheme (see above) that groups demand partners according to transactional data. By sending each call to a small grouping of demand partners, fill rate will remain high but excessive ad calls will be kept in check. The groupings are also automated and can be updated multiple times per day, depending on traffic activity.
Being smart is the key.
So many adtech functions that are devised to fix one challenge can have unforeseen consequences if we don’t take care to assess the true impact of changes like moving from waterfall ad calls to parallel ad calls. Moving forward, all adtech partners should take care to craft the most mutually beneficial and mutually considerate tools and solutions possible. Well thought out fixes save not only time and money, but also business relationships.