Ad discrepancy is the difference in statistics (i.e. impressions, clicks, conversions) between publishers and advertisers. And though it is a long-known issue it still exists and no one can find a silver bullet for it.
While the IAB terms and conditions say up to 10% discrepancy is tolerable which some may already think as a serious loss for their budgets, in some cases people experience much greater discrepancies. In this article, we will explore ad discrepancy and find out why it is much more complicated than most people believe and are there any ways to mitigate it.
A discrepancy in stats may occur due to the difference in methods and tools used to calculate analytical data and many variables (website load time, bounce rate, user interaction, using ad-blocking software) affect these calculations and make it a difficult to have one standard for all. Unfortunately, it is not the only reason for that.
We will divide the causes of ad discrepancy into the following groups:
- human factor
- technical issues
Technical issues causing ad discrepancy
Some of the technical issues cannot be easily solved and are out of our reach sometimes, but we still can try to minimise their effect. Here are the most well-known ones.
Time zone difference
While not being an actual issue, time zone difference may result in a wrong day-to-day picture in the stats. This occurs mainly because the ad delivery doesn't happen evenly throughout the day with highs and lows that may fall into different days if servers operate in different time zones.
Using different tracking approaches
Different companies may use different approaches to counting stats (impressions, clicks, conversions) and it's vital to be aware of them. For example, some may count an impression only when the webpage is completely loaded while others can use ad server requests or other event trackers. Such differences in approaches result in discrepancies in stats that cannot be fixed or mitigated. They are caused by the difference in technologies and algorithms used and just need to be considered while running your online business.
Using anti-fraud/bot or traffic quality control technologies
Any kind of traffic filtering has an immediate effect on the stats and may result in huge discrepancies. The reason for that is very simple. Everyone is free to decide what technology to use and all of them may bring different results due to different methods and algorithms used. Not to mention the case when no traffic filtering is applied.
Cross-platform and cross-device issues
Some websites may not have 100% usability when it comes to different operating systems, browsers or devices. Therefore, ads or the whole websites may not be displayed correctly and that inevitably leads to discrepancies in analytical data.
Cashed data
Users' computers, in most cases, save browsing data (history, cash, etc.) by default. While it may be useful for web surfing, saved cashed versions of websites may include ads and that proved to be detrimental for the advertising process. This way the banner first saved in cash may be displayed by the publisher numerous times while the advertiser counts only one initial impression.
Latency/poor connectivity
In digital advertising, the speed of the Internet connection really matters. Users may become annoyed by the long loading time and leave the website without even seeing the content, never mind the ads. But even if the website is super fast, users still can miss the ads due to the slow server operation. The situation when the server takes a long time to process ad requests and display ads is often caused by the physical distance between servers and users. I.e., users from Australia are likely to experience long load times while visiting European hosted websites.
Another thing that is closely related to latency is poor connectivity. Mobile devices usually suffer from that more than desktop ones as mobile operators cannot offer stable internet connection in all the places where people may use mobile Internet. Besides, some countries may have poor external connectivity by default due to internal regulations. Sometimes countries block entire IP ranges for reasons no one can predict.
These factors can add up to discrepancies and severely damage ad performance.
Large files as creatives
This issue is closely connected to the previous one. It doesn't matter how good your creatives are in case no one manages to see them due to their long loading time. It's always a good idea to use as small files as possible (considering the quality, of course) to minimise possible problems with ad discrepancy.
Human factors causing ad discrepancy
Human errors happen more often than one would probably think. The good thing is that they are the easiest to fix. Let's have a look at the most common ones.
Inappropriate ad code use
Different ad formats have different ad codes as well as the same ad formats may have different sizes. It is important not to mix everything up and use a proper ad code in every case.
Incomplete/incorrect ad code
It is essential to make sure you're using a correct and complete ad code (or URL). Otherwise, all the visits will result in no ads shown.
Please don't change the ad code or URL unless you're sure about what you're doing. Besides, always remember to check if the code (URL) is complete and all the tags and macros are in place.
Incorrect campaign settings
Running ad campaigns using various targeting options may result in ad discrepancy. Make sure all your ad campaigns settings (geo, device, frequency capping, limits, etc.) correspond to the sources of traffic used.
Conclusion
In this article, we found out that ad discrepancy has many faces and is almost impossible to completely irradicate. Not all kinds of ad discrepancy mean actual traffic loss. Sometimes it just means a different technology or algorithm. But still, it is important to understand where does it come from not to get any unpleasant surprises.