
The mobile advertising industry is based on such a shaky foundation as statistics. Fluffy - because even the base metric, the scorecard may be incorrect and not show the actual number of people who actually see the advertisement.
Advertisers who need mobile advertising believe that they only pay for the result. This may be true if we evaluate the effectiveness of mobile advertising in comparison with, say, offline advertising. In the latter, it is almost impossible to track impressions / views, not to mention other indicators. When advertising is displayed on a mobile device, almost any user action can be tracked and a detailed report compiled. Theoretically, such a system is convenient for everyone. The advertiser receives traffic and downloads, and the owner of the application - money. In practice, everything is somewhat more complicated.
The main problem is that the system of collecting statistics for mobile devices is not too accurate and reliable. The total number of ad impressions is usually greatly overestimated, since the so-called junk shows appear in the report. To solve this problem, some networks have begun to use a metric such as “mobile viewability”. It allows you to find out what percentage of advertising attracted the attention of the target audience. Ideally, the advertiser should not pay for “junk views”.
')
What is “viewability”?
Now this term is interpreted differently. On the whole, the point of view is accepted that this indicator assesses ad impressions that have been seen by the user and viewed by him. The metric is introduced so that the advertiser can be convinced of the effective use of the advertising budget.
For example, if the page is formatted incorrectly or scrolled incompletely, the advertisement may be out of the display of the mobile device. And in this case, the user will not see it, but the show will be counted and the money will be withdrawn from the advertiser for it.
Our main task in
Appodeal is to help developers squeeze out maximum advertising revenue in the application. Therefore, very often we conduct an internal examination of the data, a kind of big data. And according to our information, users do not really see about half of all the ads advertised.
Advertisers are trying to find a reliable way to move from not too effective display of advertising to provide users with views of this advertisement. To do this is quite difficult - for various reasons. But in the near future, most likely, this goal will be realized.

Traditional tracking system
Mobile advertising is automatically distributed using a number of interconnected technological processes, eventually reaching the screen of a mobile device. Typically, an advertising campaign is based on shows - when the advertiser pays, for example, 1000 pieces. Naturally, the customer wants to see the results of the advertising campaign. For this purpose, services like Google Analytics are used, which allow evaluating key campaign performance indicators.
Here it is important to understand that tracking data may contain a lot of “noise”. The report shown by statistical tools includes averaged data. The more accurate the statistics, the more valuable it is for the advertiser.
Ad impressions are not yet views.
For various reasons, certain ad impressions are useless for a marketing campaign, since not all of them provide user interaction. For example, a window can be closed even before the advertisement is fully loaded - accordingly, the user will never see the advertising content and will not click on it. But this event will be accepted by the system for viewing, and the funds will be withdrawn from the advertising campaign account.
Another consequence of this problem is that modern tracking systems lead to a significant overstatement of the cost of the campaign and give the advertiser excessive confidence in its effectiveness. If the number of views is incorrect, it can lead to the failure of the campaign objectives. To optimize resources, the advertiser must know the proportion of "fake" views and take steps to minimize the level of "noise".
Inefficiencies
For advertising material to be viewed, it must be visible on the screen by at least 50% for at least one second or two seconds (in the case of video advertising). This may seem like a simple task, but in fact not all ad networks are 100% effective due to a number of purely technical problems. Network capacity, pre-caching, and the SDK used to work with advertising materials affect the percentage of views. Also, this indicator is affected by the choice of the operating system and browser by the user.
Finally, it is worth taking into account the fact that some types of advertising are rendered better than others. For example, video ads require more time to fully load than static banners. This means that the percentage of unsuccessful impressions for video ads will be greater, since there are more chances for some kind of technical failure during loading. It is quite difficult to calculate the visibility for advertising generated by automatic services. Without new technical solutions, an attempt to use such a metric as browsing is to try to “get into the water without knowing the ford”.
We are looking for new methods
At the first stage, it is necessary to recognize that for advertisers the lack of an assessment of the perceptibility of an advertisement is wasted money. Previously, advertising network users were confident that such services provide 100% effectiveness in achieving campaign goals. In fact, this is not quite the case, as shown above.
The second stage is the coordinated actions of advertisers, which can lead to the fact that advertising networks will start to update the statistics collection methods everywhere. A new generation of software allows you to determine the exact number of views, but mobile ad networks are rather slowly updated and still use third-party tools to get statistics. At this stage, you should seek help from companies that specialize in viewability (for example,
The Mobile Majority ).
One step closer to the updated advertising market
The integration of such a metric, as visibility, into advertising networks is a necessary condition for a qualitatively new work of the mobile advertising market. Advertisements are sold in real time, and a lot of advertiser money is spent on this. Statistics based on real-world views is truly a step towards creating a reliable advertising system that will allow advertisers and publishers to make rational and effective business decisions.
Of course, the widespread use of the new metric will force a change in the approach to advertising and advertisers, and owners of applications / sites where it is demonstrated. Nevertheless, all this will allow creating a modern and highly efficient advertising services market that satisfies the needs of both advertisers and publishers.