It may come as no surprise that the Media Management, Inc. team is composed of media geeks. Media is our business, and it’s right there in our name. In addition to our media knowledge, we also consider ourselves a data first firm. Data forms the foundation of all our findings and recommendations, and our clients can attest that data is at the heart of our media audits.
Importantly, all of the data we leverage for our media audits comes directly from individual advertisers (and their agencies) and from third party media research providers. We do not believe in opaque proprietary data pools. Peter Drucker is quoted as saying “if you can’t measure it, you can’t improve it”. I tend to agree. While measurement and improvement are important, so are benchmarks that help advertisers better understand where improvement is needed most.
Here, we dive into benchmark data for digital advertisers. Benchmarks including viewability and fraud/SIVT (Sophisticated Invalid Traffic) from an industry vertical and geographic perspective will be covered, as reported by industry resource DoubleVerify. Understanding these benchmarks can help advertisers determine whether their digital media performance is aligned with industry best practices.
The Media Ratings Council (MRC) has set forth ad viewability guidelines that set a standard definition of viewability for digital ads. Standards have been established for display and video formats.
Display - at least 50% of pixels in the ad were in-view for at least 1 continuous second
Video - at least 50% of pixels in the ad were in-view and played for at least 2 continuous seconds
The MRC released a draft of new standards in 2019 that would see the pixel requirement rise to 100% for cross-media video measurement.
Not all ads that are served will satisfy the above standards. A display ad might be served to a portion of the browser window which was never in view on screen. A video ad might begin playing as a user scrolls past (without stopping) and only be in view for a fraction of a second.
This is why it is important for advertisers to not settle when receiving “served” impressions reports. A campaign could have a plan for 100M impressions, and a report could show 100% delivery. All is well and good until a savvy analyst asks about viewability. Ads served, but not viewable, will not deliver ROI.
Not every served impression will be viewable. This is to some extent the nature of digital media. The degree to which ads are viewable will vary by site and even by page based on things like the page layout, the nature of the content, and the way users interact with that specific content. Because there will always be some non-viewable impressions, it is important to understand the relationship between viewability and cost. Some advertisers only pay for viewable impressions. Some expect lower CPMs for sites with correspondingly lower viewability rates.
Now that we understand viewability and why it’s important, let’s dive into some data.
Display Viewability by Industry
The above chart shows an overall improvement in display ad viewability from 2019 to 2020, continuing a trend upward over the past several years, as the industry has increasingly focused on this issue. The results can be interpreted as a glass half-full or half-empty.
On the half-full side, viewability has improved year-over-year. If you’re an advertiser this is what you want to see. As viewability improves ads have a better opportunity to be seen by people who might purchase goods and services.
On the half-empty side, no vertical exceeds 70%, and four verticals have less than 60% viewability. Here is where advertisers understandably get frustrated.
CPG leads at 69% viewability. For CPG advertisers that means 31% of impressions were not viewable in 2020 and an astounding 45% of impressions for Telecommunications. Advertisers paying for those impressions would be better off throwing money in the street. At least that would attract consumers who have cash to spend on goods and services. This is where it is crucial to understand not only your cost structure, but your viewable cost structure with each partner.
Display Viewability by Country
We’ll now turn to display viewability by country. Out of the gate Mexico jumps out and wins the “most improved YoY” award. From only 34% viewability in 2019 to 67% in 2020 Mexico saw some great improvements. Unfortunately, that is where the good news ends.
Here in the U.S., advertisers are only recording 64% viewable impressions. While an improvement over the 59% in 2019, advertisers are still leaving 36% of impressions on the table. Recall, if the ad cannot be viewed no action can be taken that increases reach, brand affinity, sales lift, or any other metric. Again, 100% viewability may not be possible given the nature of the media channel, but advertisers who are paying for non-viewable ads or who aren’t accounting for them in their CPM evaluations are wasting cash.
Taking a more macro view, North America (U.S. and Canada) had an 8% YoY increase in display viewability, but lags LATAM (+20%) and EMEA (+14%) YoY improvements.
An esteemed American musician once said “mo money mo problems”. As advertiser budgets go increasingly digital and skew towards video, this rings true. Video CPMs, all else held equal, tend to be higher than display. This means advertisers paying for served, not viewable, video impressions are putting a larger amount of their ad dollars at risk.
Video Viewability by Industry
Looking at video viewability by industry, Retail & Restaurants manages to squeeze ahead of CPG to claim the spot for best video viewability at 72%. With the exception of Media & Sports all industries saw a YoY increase in video viewability rates in 2020.
At MMi we would consider 70%+ viewability to fall within best practices. However, plenty of room is still left for improvement with 28% of impressions not viewable for Retail & Restaurants and 39% for Automotive, Media & Sports, and Technology.
Video Viewability by Country
Advertisers in Australia showed an impressive 79% video viewability. While nearly all other countries showed improvement, India and Singapore both regressed with lower video viewability rates YoY. This is interesting as Singapore also struggled with display, flat YoY, while India did show improvement in display viewability.
Saying that ad fraud is an issue is akin to saying the Pacific Ocean is a big body of water. It isn’t incorrect, but grossly understates the truth. In 2020 alone fraud cost digital advertisers $35 Bn globally. At least $11.4 Bn of that was in the U.S. with China a distant second with $5.2 billion in lost ad spend.
As shown below, fraud rates across most industries and countries are headed in the right direction: down. However, advertisers must stay vigilant. Fraudsters are smart, and they understand opportunity cost better than most. With CTV averaging $20 CPMs the bad actors follow the money. DoubleVerify reported a 161% YoY increase in CTV fraudulent traffic rates for Q1 2020 vs Q1 2019.
Fraud by Industry
Let’s start with the positive: all industries included in the data set saw a reduction in fraud YoY. It will be interesting to view the 2021 data next year to see if the trend holds. Given the COVID-19 pandemic, reduced fraud could be a result of advertisers having an extra careful eye on their partners. Better fraud blocking solutions could also be a contributing factor. Either way, the decrease in fraud for 2020 should be considered a big win as more people were spending a large portion of their time online with increased ad dollars following.
Improvement across all industries is still needed. MMi considers < 1% of impressions lost to fraud/SIVT as a best practice. CPG comes in at 1.5% for the lead whereas Media & Sports is showing 3%. All other categories range between the two with a median of 2.10%.
Fraud by Country
With the exception of Singapore, most countries included in the data set saw a reduction in ad fraud. Australia, Canada and Germany all remained flat YoY. Of note is India’s large decrease going from 2.8% in 2019 to 0.6% in 2020.
Regions varied in their success with fraud on different devices. North America performed best on desktop (-42% YoY) while APAC saw +33% YoY. While APAC finished last in the desktop category, they dominated the mobile app environment with a -85% YoY reduction in fraud rate. EMEA was the laggard in mobile app with a -27% YoY change in the fraud rate. Finally, LATAM led in mobile web (-36% YoY) while APAC lagged behind (+4% YoY).
Some takeaways from 2020 include:
Viewability (display and video) has improved YoY across industries and countries with a few exceptions
Fraud rates declined across industries and in most countries, but…
Advertisers need to remain vigilant as bad actors follow the dollars to CTV
Advertisers looking to improve viewability and decrease fraud rates in 2021 should ensure they are working with a reputable verification partner, if one is not already in place. Best practice would ensure that all agency and vendor agreements clearly state the provider to be used.
Include contractual language clearly stating the advertiser will not pay for impressions crossing an advertiser-specified threshold for viewability and fraud. Viewability of 70%+ and fraud < 1% of impressions is a good starting point.
A final consideration would be conducting a media partner review. With the deprecation of third-party cookies looming, now is an ideal time to discuss new targeting strategies for a post-cookie world. This can, and should, include a review of media partners to assess any targeting solutions they provide and how they address viewability and fraud on their properties.
If you have concerns about viewability and fraud in your digital campaigns, or just want an independent review to validate agency reports, get in touch with MMi today.