What Is Ad Discrepancy and How to Minimize It Today and in the Long Run [+Checklist]
There is already so much to memorize in programmatic advertising: from RTB and waterfalls to pricing models and optimization techniques. Still, even with all the mightiness of the automated ad campaigns, we have to double-check the analytics and reporting because of the digital advertising discrepancies. What? Why? Can we do anything about it?
So, you launch a new advertising campaign. The sun is shining, the forecasted numbers are promising you sky-high CTRs, 110% viewability, and a level of engagement Bella Hadid's swimsuit pictures have never seen. Then, the end of the month arrives and the analytics from the sell-side comes in.
You look at them, then at the forecast, and you cannot help but notice a difference in the numbers. This, my friends, is called ad discrepancy. It is this difference in analytics that makes advertisers suspect publishers of being untrustworthy and vice versa.
In reality, ad campaign discrepancy is kind of a norm in the ad tech world, depending on the %, of course. Also, there are a number of ways to minimize these metric differences very quickly. But first, we would have to understand what ad discrepancy is.
What Is Ad Discrepancy and How Is It Calculated?
It is said that an image is worth a thousand words, but sometimes it is still better to stick with the words, especially when it comes to media buying.
Definition of Ad Discrepancy
Ad Discrepancy is a difference in metric calculations between the sell and the buy-side at the end of the reporting period. While the data discrepancy can be traced in various metrics, for example, time spent on the page or conversions, the most frequently affected are clicks and impressions.
A number of factors may cause a discrepancy in an ad server: from the time difference and site latency to the difference in the algorithm operations. Some of these factors are easily managed, while others may take time to pinpoint and therefore eliminate.
How to Calculate Ad Discrepancy
If you cannot spot the difference with a "naked eye," then here is a user-friendly online calculator link that can spot the discrepancy for you. However, the classic formula to calculate ad discrepancy (regardless of the metric) is:
- Subtract the number you received from the other party (be it sell or buy)
- Divide it by the other party's number and multiply by 100.
- The number you receive is the % of the discrepancy.
Important note:: this number does not necessarily indicate that the other party is wrong or untrustworthy. It means that both of you have differences in measurements, and you need to look into them to minimize this number in the future.
What is the Acceptable % for The Ad Discrepancy?
Everybody wants their end-of-month reports to state 0% discrepancy, based on the previous data and numerous accounts, but unfortunately, this is Non-Possibile (please read with a strong Italian accent).
IAB's Standard Norm and Conditions for Digital Advertisers states, quote: "If the difference between the Controlling Measurement and the other measurement exceeds 10% over the invoice period and the Controlling Measurement is lower, the parties will facilitate a reconciliation effort between Media Company and Third-Party Ad Server measurements." Otherwise, the matter should be resolved privately, depending on the situation.
So, to put it in simple words: if the ad impression discrepancy in programmatic advertising is 10% or less, there might be room for improvement but not for worry.
8 Most Common Causes of Ad Discrepancy [with Solutions]
Factors behind ad discrepancy may vary but mainly it boils down to the difference in measurements/settings or human inattentiveness. Below are the ones that can be managed quickly and ones that might need longer to solve.
Not So Quick to Solve Causes of Ad Discrepancies
Ad discrepancy reason #1: Different measurement systems
For instance, we have three parties involved in our ad campaign (sell-side, buy-side, and third-party analytics service), and our main task is to, say, track impressions. However, one system may record an impression when a user only requests the page, the second - when the user loads the page but not the ads, and the third one - only when the user loads the page and all the ads.
On top of that, these parties can have different measurement systems. For example, Google Ads filters all the invalid traffic while other analytical systems (like Google Analytics) show all the data. Also, there could be way more than three parties involved (from DSPs and SSPs to ad exchanges), and each of the parties can have a different measurement system or preset metrics.
You see, none of them would be wrong or dishonest - they just simply use different ways of measurement.
Solution: Always discuss the metrics and the measurements of all parties involved before the campaign is launched. This way, you and your partners will be on the same page and have the same statistics in your reports. It seems to be better to ask them out for a coffee break instead of that discrepancy calculation session, doesn't it?
We also suggest that you check the results daily to track any signs of digital advertising discrepancy at the earliest stages. The latter, however, will depend on your human resources.
If you decide to involve any third-party platforms for ad campaign attribution, it's best to choose the same system your partners already use.
Ad discrepancy reason #2: Ad blockers
According to Hootsuite, around 40% of onliners use ad blockers globally. If you are not familiar with the term, an ad blocker is a technology that blocks ad requests, resulting in the user not seeing the ad. While most systems can track when the ad blocker is used, some cannot. This alone may lead to significant ad discrepancies when one party will see and record a user on the site/page while the other will not count the impression.
Solution: Just like with the difference in measurement systems, here you would have to find out how the other parties are measuring their impressions.
Ad discrepancy reason #3: Invalid traffic
Bot farms, domain spoofing, and similar malware intend to fool the system and get money from the advertisers. Luckily most trustworthy ad servers and DSP providers have special tracking tools and security systems to recognize malicious traffic and prevent it from causing harm.
Moreover, when you have too much bot traffic, Google might block you, and it will not be very easy to get back into the online advertising game.
Solution: Work with dedicated tech partners. For example, Epom's Ad Server uses bot filtering at an engine level: you can check the quality of the inventory in your account.
Ad discrepancy reason #4: Time difference
If the advertiser and the publisher (and any other third parties involved) work in different time zones: ETS and GTC, for example, there will always be a time gap giving room for the discrepancy: when someone has counted impressions/clicks, and another hasn't.
Solution: In this case, you either agree on the same time, like 7 PM to 7 AM, using the time zone of one of the parties, or you can go with the UTC zone - Coordinated Universal Time. Also, highly recommended to do this before the launch of the ad campaign.
Quick(er) to Solve Causes of Ad Discrepancies
Sometimes you have to get in touch with your ad tech partners in order to align your metrics, time zones, and also by agreeing on ad blockers, and other times improving ad discrepancy percentage is just a matter of a few clicks and putting your creatives on a healthy diet. Below are the latter ones.
Ad discrepancy reason #5: Site latency and heavy creatives
If the site is not loading for 2-3 seconds, the bounce rate increases dramatically (in 2017, Google mentioned something like 32% after 3 secs). While it seems that it can only affect the site's performance, it can also cause discrepancies in ad calculations. One party might count a user visiting the site as an impression when in reality, the user has left before ads, and content even appeared on the screen.
We want to single out two factors that may influence site latency: heavyweight creatives and the ad server's slow performance due to geolocation. If an ad server is in the US, the user in Europe or Asia might experience a slight delay in the site's performance, which might eventually affect the analytical numbers.
Solution: We agree that a catchy creative is always a good idea, however, the site's loading speed is more important. So, be sure to reduce your creatives' size to an acceptable one.
If you are using an ad server of a DSP then you can always run a test.
The same goes with the latency - if you use an ad server with server capacity on different continents, then the latency may be reduced to milliseconds.
Ad discrepancy reason #6: The difference in ad campaign settings
This ad discrepancy mainly happens when working with direct deals, when the advertiser communicates directly with the publisher. A mixture of human error and differences in measurements may happen when parties try to agree on the same metrics: capping, geos, limits, etc. but eventually go out of sync.
Solution: Set up limits, capping, and targeting with your partner/s, decide on the launch and the end of the advertising campaign (remember the time zone, heavy creative, and all the aforementioned).
Ad discrepancy reason #7: Missing ad tags
Be sure to match the ad tags both on the publisher's and ad server's side. Another issue may occur when your ad tag is displayed in a specific country, and the statistics are gathered globally.
Solution: Be very attentive with ad tags; double-check if you copied the whole thing. In case you missed something and an ad tag is not working, issue/ask to issue a new ad tag and insert it very carefully ("Handle with care", "This side up" kind of careful). For VAST tags you can always use a special feature that will double-check everything for you.
Ad discrepancy reason #8: Missing macros
Partially set macros or deletion of some parameters from the tag, incorrectly set pixel tracking, or passbacks will interfere with sending dynamic data back to your ad partner, leading to a poorer targeting, analysis, and as you probably have guessed by now... ad campaign discrepancy.
Solution: Ensure that all the parameters that the advertiser requests are clear to you and also be sure that your server has suitable macros for transferring these parameters.
Epom ad server has a broad base of supported macros, an opportunity to implement pixel tracking or passback/piggyback. If a specific parameter needs to be passed, it can be passed through the custom parameter. Thus you can fill any gaps in the absence of macros.
How to Deal with Ad Discrepancy: Key Takeaways
Now, gather around, and let's see what we have learned today.
- DEFINITION: Ad discrepancy is an inevitable side effect that comes with the blessings of digital advertising. However, it should not be seen as doom, since as stated by IAB, anything under 10% is acceptable.
- ACCEPTABLE %: Even though minimal discrepancy in ad calculations such as 10% is a norm, a higher number may cause damage to your final reports.
- CAUSES: A number of factors contribute to a higher % of ad discrepancies, such as differences in measurement systems, time difference, heavy creatives, etc., and the good news is that we can work with all of them.
- SOLUTION: Choosing a trustworthy and experienced tech partner is one of the key solutions to any digital ad problem, including ad discrepancy.
Bonus: the "Ad Discrepancy Checklist" to Help You Juggle All This Digital Ad Drama:
Want to test-drive Epom Ad Server for free and push your ad discrepancies to the lowest possible?Sign Me Up