Just recently, AerServ released a new version of our monetization analytics, giving publishers a deeper view into the flow of ad events within the app. You can read more about the new metrics here. We thought it would be helpful to follow up the release with more information about these reports, and some best practices for using the new information.

A Background 

Within a desktop environment, or when mobile ads are integrated via tags or API, the flow of events is very simple. This flow, depicted below, starts with the opportunity to show a user an ad, which is the initial call to the ad server for an ad. As soon as this happens, the ad server will send out ad requests to all demand sources assigned to the placement. With mediation, this will be a one to many relationship – one ad opportunity will result in multiple ad requests, as multiple buyers will be sent requests for ads. Once a valid ad is selected by the ad server (after filtering and decisioning logic do their thing), the ad is delivered to the user, resulting in an ad impression. Simple 1, 2, 3, of events.

The New Metrics

Within mobile, and with SDKs, we have a number of other events available to us that provide valuable insights. As many mobile app developers know, with mobile advertising we have the wonderful luxury of being able to pre-cache an ad. This process, part of the diagram shown below, runs in the background, unbeknownst to the user. When a user clicks the app icon on their phone screen, and the app is launching, it can also initialize the AerServ SDK and start the ad call process way in advance.

Why do this? The goal of this process is to reduce latency and improve the user experience by having the ad stored on the device and ready to be presented to the user when they are at the point in the app when we can show them an ad. One of the large complaints from users, and reason for ad blocking, is due to latency – pre-caching helps.

When the first call is made to the app, we log an Ad Opportunity metric. Immediately, the SDK and ad server go to work and request ads from all buyers, resulting in multiple Preload Requests (same one to many relationship as above). Once a valid ad is selected by the ad server (after filtering and decisioning logic do their thing), the ad is stored on the device, resulting in a Preload Ready event. This means we have an ad ready to go! But, it doesn’t mean that a user is ready to view the ad. When the user arrives at that magical point in the app where we can show them an ad (ie: transition between levels, rewarded video placement, et), the app invokes the Show Ad Attempt which fires the event signifying that we tried to show the user an ad. If the ad was able to be retrieved and shown to the user, then the ad impression event is fired. Below is how the events appear in the AerServ platform.

What to do with User Ad Rate?

“What the heck is user ad rate, anyway?”  This metric shows you the percentage of time that a user made it to the point in the app where an ad could be shown. The calculation used is the total number of show ad attempts divided by the total number of ad opportunities. For example, if you had 100 people click the icon to start the app, but only 25 of them clicked the “watch video” button for your rewarded video offering – you have a 25% user ad rate. You want this number to be as high as possible to maximize your revenue potential. If I see a 25% user ad rate, it means I have 75% room for improvement.

If the user ad rate is low, it could signal trouble with the location of the ad, or how it is being offered to users. Is it buried deep in the app, or hidden behind in-app purchase promos? We’ve seen these scenarios with publishers, and the low ad rate is no surprise. Getting the user ad rate as close to 100% as possible means you’re monetizing more of your user base, and we advise our customers to optimize their ad placement strategy.

What to do with Ad Use Rate?

“Ok, but what the heck is ad use rate?”  This metric, sounding too similar to the one above, shows you the percentage of time that a valid ad ended up being delivered, or viewed by a user. The calculation used is total number of delivered impressions divided by the total number of preload ready events. For example, if we had 100 preload ready events fire, that means we had 100 ads locked and loaded, ready to show a user. If we only had 50 impressions delivered, that means our ad use rate was 50%, and that we had 50% more ads available to show.

This metric is an indicator of opportunity – in the above example it shows that we could be generating 50% more revenue if we optimize, tweak, and adjust our game. The availability of ads isn’t holding us back – the presentation and user flow is.

This metric could also be an indication of a problematic buyer or creative. If we see a low ad use rate, it can be helpful to dig deeper and view the report by ad source. When the ad use rate is disproportionately low for a particular ad source when compared to other ad sources, it is recommended to discuss it with the buyer. It could be a bad creative coming through, or a technical bug.

What’s next?

We hope that you’re enjoying these new metrics, and they’re helping you to better understand the ad flow, and optimize your monetization. Within the next few months, we’ll be expanding these reports to include additional features and filters for increased granularity.