HOW AD VIEWABILITY IMPACTS ADVERTISING REVENUE

In the past, I have written about the relationship between rising ad blocker adoption rates and declining publisher revenue. I concluded that there is a direct relationship between a publisher’s reader ad block adoption rate and their lost advertising revenue. In short, Lost Revenue = Current Revenue * (a/1-a) where ‘a’ is the ad block adoption rate.

 

Recently I have been researching whether a similar relationship exists between a publisher’s revenue and their viewability rate, and through my testing, I found that an extra second of ‘in-view time’ means extra $0.15 eCPM on average.

 

Let me show you how:

 

The definition of “viewable” for an ad, and the idea behind it is pretty simple - advertisers really only want to pay for ads that are actually seen by human eyeballs. Per the standard definition, a display ad is viewable if at least 50% of its pixels are on screen, in the viewport, for at least 1 second, and a video ad is viewable if 50% of its pixels are on screen for at least 2 seconds. If an ad impression meets these criteria, the advertiser is satisfied and the publisher gets paid.

 

If you go by this standard for viewability, you'll notice that the placement of the ad on the page, whether it appears above the fold or below the fold (ATF vs BTF), isn’t taken into account. However, ATF (above the fold) units command a premium because they often have higher viewability rates than BTF (below the fold) inventory. This makes sense because the ‘viewability clock’ for ATF inventory starts counting immediately when the page loads, but BTF inventory doesn’t start counting until the reader scrolls down far enough for its pixels to start coming into view. Even though BTF inventory has much lower viewability rates than ATF inventory, multiple studies show that it is still is able to manage around a 40% viewability rate. So if we know that ATF inventory is on screen longer than BTF inventory, and therefore has higher viewability rates and drives higher CPMs, we can assume there is a relationship between viewability, time in view, and CPMs.

 

My hypothesis is two-fold: 1) Higher viewability rates command higher CPMS, and 2) Viewability directly correlates with ‘time in view’. The longer an ad is on screen, the more it is likely to be counted as viewable, and the higher the CPM it will command.

 

Given this hypothesis, let's break inventory into two groups - ‘0-1 second’ time in view versus ‘greater than 1-second’ time in view.

 

Impression volume % vs. Exposure Time (Sec)

 

The above chart represents an average US-based publisher for their display inventory (US publishers average around 56% viewability rate)

 

Let's now assign the CPM/RPM to each bucket (assuming a comScore top 150 property) which typically is:

 

CPM/RPM applied to each bucket

 

 

That's right. You get paid the floor, which is usually $0.1 CPM, for the ‘0-1 second’ time in view inventory, but get 12 times that for the ‘greater than 1-second’ inventory, averaging around $1.20 CPM. Your exact results and CPMs may vary, but the key point is that if you maintain a roughly 50% viewable rate, then half of your ad inventory is not making you (much) money and the other half is where you make almost all of your money.

 

For the largest publishers, the relationship between a publisher’s revenue and their viewability rate is:

 

Revenue Composition

 

In short, you don’t really get paid for non-viewable inventory.

 

What is interesting is that there is nothing magical about viewability being defined as 1-second. Many groups are talking about viewability being redefined as 2 or 5 or 10 seconds in view in the future. If we continue our analysis with a focus on ‘time in view’ the results get very compelling at the time in view of 10 seconds:

 

Time in View

 

On average, US publishers should see around 32% of their total inventory in view for 10 seconds or more, but this inventory will generate 63% of the total revenue. It gets paid 6X the rate of the 0-10 seconds time in view bucket. Time in view matters a lot!

 

From here then, it is very easy to see that CPM tracks closely with time in view - roughly 15 cents of additional CPM are gained for every additional 1 second in view.

 

CPM vs Exposure Time

 

Based on the prior analysis we can safely say that if someone improves their viewability rate they should improve their revenue. For example, adding 10% extra viewability means that the additional revenue is equal to: the volume of impressions implied by the 10% multiplied by the CPM delta between 0-1 and > 1-second buckets.

 

Added revenue = Current impressions x  ΔViewability Rate x ΔCPM

 

Where ΔViewability Rate = the change in viewability rate (56% to 60% = 4% change)

 

And ΔCPM = current CPM minus your floor CPM

 

An extra second of in-view time means roughly an extra $0.15 eCPM.

 

To conclude - I have found that a relationship does exist between viewability and revenue. Publishers looking to add additional revenue should be attacking their viewability rates just as aggressively as they are their ad blocker adoption rates.