Facebook’s Delivery Insights gives mobile advertisers the bigger picture over ad set performance. It offers a place for mobile app marketers to keep an eye on ad performance changes, and in particular helps them understand what could have caused them in order to optimize accordingly. Delivery Insights is a pretty powerful feature in its own right, and the large amounts of data it provides can be taken to new heights when merged with the right kind of AI.

Features to Note

Audience Overlap

Facebook prevents you from bidding against yourself by absorbing your campaign data and removing most ad sets (in the same account) with a high ‘Auction Overlap Rate’ from relevant auctions. These ad sets most likely will suffer from weaker performance as they will not have a chance to be delivered. Delivery Insights also offers a view of ‘Most Overlapping Ad Sets’ – identifying the ads (indicated by their ad set IDs) that are causing auction removal. This data allows users to conveniently view first, second and third most overlapping ad sets. When overlapping starts negatively impacting ad results, Facebook advises to start consolidating, yet when this solution doesn’t quite fit, it could be time for other solutions to be brought to the table.

Audience Saturation

Audience Saturation is the moment where performance starts to drop as ad frequency rises. This data set allows advertisers to carefully examine their desired outcomes and adjust any elements accordingly. Facebook added this Audience Saturation measurement into Delivery Insights to allow users to understand: how much of an audience an ad set has reached (‘Audience Reached Ratio’) and what percentage of an audience has already been reached (‘First Time Impression Ratio’); the latter is calculated as a percentage of impressions an ad set received in the past day.

 

How To Use This Data

The latest changes to Facebook Delivery Insights allows advertisers to be proactive in investigating campaigns that display sudden shifts in performance. It’s safe to say that this data, when used correctly, could make or break a User Acquisition campaign as it has the power to:

Reduce CPI/ Increase ROAS: Optimizing campaigns is notably the main purpose of Delivery Insights. The latest Audience Overlap and Saturation features allow users to take a deeper dive into ad data to assess targeting, relevance, delivery, budget pacing, and overall performance. For example if an overlap rate is quite high (10%+), ad sets are figuratively stuck and need to be set free. Overlap data can be used to discover and wipe out the root cause (which might even be a poorer performing ad set) which will in turn allow CPIs to drop and ROAS to increase.

Improve Delivery: Ad set overlapping can also stall delivery. This new data, when interpreted correctly, can provide valuable insights to offer stronger ad delivery. One way this can be achieved is by finding the winning elements in overlapping ad sets and merging them together to create one ‘super-ad set’ with a larger budget. Overlap and Saturation results also allow app marketers to compare and contrast information to ensure that the right ad set is being delivered to the right audience via campaign objective. If two separate ad sets are in play with two different objectives, reducing audience overlap would be essential for stronger delivery.

These results can also point to a need for both creative and audience refreshment. With Delivery Insights latest features, advertisers have a larger pool of data to play with, which can help obtain insights on when and where to refresh creatives and when new audiences are needed, post saturation point.

Adding AI

App marketers everywhere are welcoming the more recent changes to Facebook Insights, as the feature allows them to utilize data to make optimization decisions on their campaigns in an easier, more comfortable way. Applying AI technology to Delivery Insight data should be near top of every mobile app marketers to-do list, hence why Bidalgo is already providing full Insight integration into their own platform, accessible to all Bidalgo users. The Bidalgo ‘Overtime Graph’ for example, offers an essential way for mobile app advertisers to easily visualize overlap KPIs and how they impact business results. The graph takes Facebook overlap data and delivers advanced functionality, allowing users to breakdown and compare overlap metrics to any predefined KPI from ROAS to CPI for ultimate optimization. With this graph in play, the user can make more accurate decisions and can more clearly understand how certain actions impact overlap behavior over a set period of time.

Delivery Insights can be an extremely valuable tool for mobile app advertisers… if you have the knowledge and know-how to deal with the large amounts of data it yields. It is a remarkable step in the right direction for Facebook’s backend, with more exciting features expected to arrive throughout 2018. However, using these new data sets to optimize manually can prove a challenge. Hard and fast rules no longer apply and app marketers often find themselves in the situation where they need to quickly establish winning and losing creative assets in order to move forward quickly. This is most efficiently done with AI platforms that have the ability to suss out winning elements in moments and can use Delivery Insight data to really put ads to work.

Infusing the correct AI elements into Facebook Delivery Insights can help mobile app marketers make sharper decisions not only based on conversion and impression data for auctions, but also Facebook auction competition levels, audiences, saturation and overlap entities. When viewed in a multichannel environment, these results can tell many different stories and have the power to significantly improve performance.

REMEMBER!

  • In order to use Delivery Insights, ads must have have at least 500 impressions spread over 5 consecutive days.
  • Delivery Insight data will only be accessible from current day -1. Data from ‘today’ will render ‘N/A’