In numerous “advertising attention” studies now done across Facebook, YouTube, Instagram and TV in Australia and international markets, Professor Karen Nelson-Field has data which answers the critical question many marketers and media specialists are either still not asking or care for: How much time does an ad need in front of a user or viewer to be effective? On average around 50% of “time-on-screen” has no viewer attention paid at all. And no platform is immune. Here’s everything you need to know – fast.

Last week a billion dollar-plus question was posed by Jonathan Fox, Director of Effectiveness at Nine Network, who asked Mi3 readers a question that everyone really wants to ask Facebook but don’t have a direct email to do it: “Is 1.7 seconds enough for an ad to be effective?”
The short answer is, no. But the longer answer is far more interesting and, perhaps, surprising. And it goes far and beyond Facebook.

Here’s why:

• The adequate number of seconds is not a question specific to Facebook, it’s a question for the entire media and advertising system
• “Attentive” seconds as a measure can truly answer the question.
• Beware proxy measures of attention or engagement – opaque metrics of time on-screen are not what they seem and can represent viewer or user distraction, just as easily as they can engagement.
• On average around 50% of viewer or user time on screen has no attention paid at all.
• Get this: Better performing media platforms (in terms of overall sales uplift), seem to perform better at lower numbers of attentive seconds than poorer performing platforms do even at higher numbers of attentive seconds.
• The simple truth is you should pay more for quality attention.

You know what it’s like reading an article and talking (sometimes yelling) at your screen – we all do it. You might also know that feeling when you realise you can actually answer a difficult question. This time I thought I’d come forward and talk to your screen, not just mine.

The extensive multi-country work we have done on attention and the effectiveness of the (low) global viewability standard, I’m asked this question often: Is the Media Ratings Council’s international benchmark of 50% ad viewability and a two-second cut-off enough?

We do see a material uplift in sales and attention when both pixels – or the actual size of the ad on a screen – and time, are increased. We also know that pixels (size of the ad) are more important than time. There is no point in an ad being on the screen for longer, if it can barely be seen.

I have published this data before. What I haven’t talked about is the findings from our multi-country studies that specifically describe differences in effectiveness between groups of ads that were seen at above – and below – two seconds of time.

Let’s just get a few things straight first.

  1. My answer to this question has no relation to the Facebook boycott, which is a totally separate (and important) debate.
  2. This answer is not directed at Facebook; this is clearly a marketing, media and advertising ecosystem issue.
  3. We look at the number of “attentive” seconds, not the number of seconds an ad was on screen, and there is a big difference.

Attentive seconds can truly answer the question being asked.
As a measure it represents, on average, how many seconds of human attention were actually paid to advertising. Time-on-screen, does not. We see in our data that time-on-screen, a metric that is used as a proxy for attention, in reality tells us little about whether a human has actually seen the ad or not.

We find that on average around 50% of time-on-screen has no attention paid at all – so time-on-screen can equally represent distraction as it can engagement. And before you start yelling at your screen, no platform is immune to this problem. My point here is, that if we truly want to answer this question then we need to consider ads that were actually looked at above and below two-seconds.

The data presented below comes from one of our international collections (pre-Covid). It represents around 4,500 single ad views over four platforms: Facebook, YouTube, Instagram and Linear TV. For this collection, we also collected brand choice after the ad exposure in order to determine Short Term Advertising Strength, a peer-reviewed, globally-accepted measurement benchmark (STAS). STAS, at its most basic, describes sales uplift after an ad was viewed. We use it to cross-check the relationship between attention and meaningful behavioural outcomes.
Quite simply, this table shows the difference between the STAS above and below the two-second mark across the four platforms.

There are a few things of note here.
First, it shows conclusively that there is a difference between above and below two seconds, so Jonathan Fox is right: two seconds is not enough when you consider the money left on the table (in terms of sales) when an ad is viewed for longer.

Second, and this is the surprising bit, you will notice the relative similarity in the number of STAS points difference across the platforms (no significant deviations from the average). What this says is that two seconds is not enough regardless of platform, even TV, and the size of the effect is similar.
Third, better performing platforms (in terms of overall STAS), seem to perform better at lower numbers of attentive seconds than poorer performing platforms do even at higher numbers of attentive seconds.

This is not super surprising to me. It comes back to the nature of the viewing experience. Platforms like socials, where the user has a greater opportunity to not look at the ad simply due to the user experience, suffer the consequences. Some platform designs simply create an environment with more distraction and lower ad visibility.
So why does this matter to you?
If someone is telling you that two seconds is enough, don’t believe them. A deeper understanding of real human attention within the advertising ecosystem, is what it takes to leverage ROI for media spend.

Don’t mistake this comment for advocacy of lower CPMs though. I have said this before and will say it again: seeking low cost reach is a fool’s errand. The simple truth is you should pay more for quality attention if you want to grow your brand.

Source: https://www.mi-3.com.au


No matter what data source you consult, the flight of TV viewers from traditional linear TV to Netflix, Roku, Hulu, Disney Plus and any number of other VOD platforms is unmistakable. Consider that 40% of adults in U.S. TV households are now watching video on a TV set via a connected device on a daily basis, compared to just 1% in 2010.

Reports are circulating that streaming to TVs has more than doubled during the pandemic because people are spending more time at home. Live sports’ prolonged absence is also a factor here because sports have historically been the main reason not to cut the cord for many cable subscribers. (It remains to be seen what will happen now that sports are returning.)

But viewership trends don’t always neatly track to ad spending, and a Connected TV (CTV) dollar simply isn’t the same as a linear one. (To be clear, CTV is a catch-all phrase for systems that allow people to watch video content via a streaming internet connection. It includes built-in Smart TV interfaces, stand-alone streaming devices such as Roku and Chromecast, connected video game systems and Blu-ray players.)

The question now is what this surge in CTV consumption means for advertising supply and inventory value across linear and VOD channels alike. In my view, the impact of CTV’s surging viewership on linear TV’s standing will be blunted by the lack of maturity and trustworthiness of VOD platforms.

Beware of CTV Ad Fraud

Fraud plagues all digital platforms, and CTV is no exception to that. In fact, what may have been the biggest CTV scam of all time came to light recently. In January, a bot network called ICEBUCKET impersonated more than 2 million people in over 30 countries, generating around 1.9 billion ad requests per day—or 28% of total volume—on programmatic platforms for CTV. According to a White Ops investigation, ICEBUCKET counterfeited more than 300 publishers, tricking advertisers into thinking that real people were seeing their ads.

This kind of fraud doesn’t happen on reliable and brand-safe linear TV. If advertisers can’t be sure they’re getting true viewers from CTV’s growing ad supply, they’ll refuse to allocate additional spending to ad-supported video on demand (AVOD) channels—and for good reason.

Conduct an ROI Analysis

It’s also important to remember that CTV is not yet a very effective vehicle for mass reach. For that reason, it’s still of limited interest to global consumer brands, such as Coke and Pepsi, who focus on brand awareness and affinity, not “down funnel” metrics that are supported by CTV’s targeting capabilities. For the most part, AVOD services don’t offer mass reach, and platforms such as Roku and Hulu with larger audiences have significantly higher average CPMs than linear TV. As a result, traditional linear TV still offers the best value for advertisers who care most about tonnage.

As part of the digital advertising ecosystem, with its array of intermediate technologies, CTV imposes a “tax,” which eats into sellers’ net revenue and profit margin. For example, a large amount of CTV inventory is sold in programmatic marketplaces, which take a percentage of the value of every impression sold. Sellers should conduct an ROI analysis to determine whether the higher CPMs from CTV actually yield a better return than linear after accounting for the additional costs and human resources required to plan and deliver CTV.

For “up funnel” marketers, I believe that CTV is most valuable in the short run as a reach extension for linear TV. This is especially true in the COVID era, when big brands and local advertisers alike need sophisticated targeting capabilities to tailor content based on how coronavirus infections are trending and how the economy is impacted on a local basis. After all, restaurant commercials would land badly in cities where restaurants have been ordered to close—as would soda commercials depicting large-group gatherings.

Source: https://blog.furiouscorp.com