The latest edition of the IAS Media Quality report for H2 2017 featured U.S. display and video fraud benchmarks as well as global mobile fraud benchmarks. Advertisers have been particularly concerned about fraud over the last year, and those concerns were reflected in many of the questions we received during our webinar Transparency in digital: Metrics for the strategic marketer.
Please find a few questions with answers below that we were unable to address during the MQR Webinar.
Q: Why are fraud benchmarks for publisher direct generally lower when compared to programmatic?
The answer is largely related to inventory sources. Publisher-direct buys tend to be composed of premium inventory. Premium publishers are more likely to rely more heavily on a managed services buying model and less on placing inventory in an open exchange. These publishers have more safeguards in place and are less likely to attract high levels of fraud compared to lower-tier publishers.
Programmatic exchanges tend to contain a more heterogeneous collection of inventory. While there is certainly premium inventory to be found, particularly in private marketplaces (PMPs); however, open exchanges also often include non-premium inventory. Some of these publishers have less robust fraud control measures in place than their premium counterparts.
Additionally, programmatic buyers have traded some transparency for efficiency when purchasing inventory. This is why it’s so important to use pre-bid solutions to regain control over media quality and ensure your programmatic buys are protected against fraud.
Q: When it comes to non-optimized fraud, why might we see higher rates of ad fraud for publisher direct buys than in programmatic buys?
Non-optimized fraud rates are benchmarks indicating what a typical advertiser can expect if there is no fraud mitigation strategy in place. A mitigation strategy could be as comprehensive as leveraging a 3rd-party verification and optimization partner, or as simple as buying high-quality inventory with lower rates of fraud.
Conversely, non-optimized fraud rates are based on inventory where the advertiser has leveraged no concerted strategy to mitigate fraud. Publisher direct inventory is typically higher in quality and therefore more likely to have lower fraud rates due to the ability of advertisers to select low-fraud environments. Since non-optimized fraud excludes campaigns with a fraud mitigation strategy, the inventory is likely to be of lower quality and include environments with high levels of fraud. This applies to non-optimized fraud benchmarks in both publisher direct and programmatic buy types. Therefore, fraud rates within publisher direct buys can equal that found in programmatic or even exceed them.
Q: Can you address prevention vs. detection, and where the industry currently sits with respect to this?
Fraud detection is an important first step. In order to strategically tackle any challenge, you must first comprehensively understand the nature and scale of the problem. Through measurement and reporting, advertisers are able to better size up the problem.
Measurement and reporting allow advertisers to compare the amount of fraud they incur from various sources–between publishers, on an open exchange vs. PMP, from managed service buys vs. programmatic, and so on. Once these sources are identified and better understood, a strategic prevention strategy can be implemented.
The best fraud mitigation strategy utilizes a multi-faceted approach:
- Direct publisher outreach to tactically address fraud
- Strategic buying so advertisers prioritize inventory with low-fraud levels
- 3rd-party blocking capabilities to ensure ads are not being served to known sources of fraud
- Pre-bid optimization technology that allows advertisers to avoid bidding on impressions sourced from fraudulent origins
Q: Do you have fraud benchmarks across global markets? If so, are there major differences between them? What accounts for those differences?
We have fraud benchmarks from over 30 global markets. Desktop display is the best fraud benchmark to compare globally as many emerging markets are not as advanced in terms of mobile advertising. This impacts the amount of data we are able to collect within those markets.
Looking at desktop display, we do see a fairly wide variance in terms of fraud. For example, the global “optimized against fraud” average for desktop display is 0.7% compared to 1.8% for the U.S. and <0.1% for Thailand. So, why is the U.S., an advanced market, so much higher than the global average?
Fraudsters’ objective is to make money. As a result, fraud follows opportunities for revenue. In advanced markets, there is simply more opportunity to earn: transaction volumes are higher, CPMs are elevated, and there is more advertising, in general, to go after. This is why it’s important to ensure you have a global anti-fraud solution that can measure and reduce fraud across all markets within which your business operates.
Have more questions? Check our our other H2 2017 Media Quality Report Q&A’s below: