Pathlabs

View Original

Ad Impressions That Matter: Using Viewability Benchmarks

In the early 2010s, the industry as a whole — brands, agencies, publishers, etc. — realized we all had a big problem: advertisers were buying ads that no one could see.

So, they all got together in 2014/2015, created some standards. Now viewability, the measurement of viewable impressions, is a huge factor in modern campaigns

Viewability’s impact has been massive and to understand how to properly leverage it, we first need to dive into the basics of what it is and how we got here.

Click here to jump down to our 3 Viewability Benchmarks.

See this content in the original post

What is Viewability?

Viewability is a metric that tracks impressions that users can see.

But wait, aren’t all impressions seen by users? Not necessarily. Let’s hop into our time machine and travel back to the mid-90s.

When the very first ad servers were just getting off the ground, they made one big, but understandable, assumption: when a publisher requests an ad, that ad would always be visible to the viewer on a web page. After all, why would you sell space for an ad if no one ever sees it?

There are various reasons this assumption was and still is wrong, like technical limitations, publishers gaming the system, and more. The result was that advertisers were paying for ads that were never seen — like a display ad loaded at the bottom of a page, which the user never scrolls down to.

Things were getting pretty bad for advertisers. In their article on viewability, Digiday says,

[In 2014], depending on whom you believe, up to 54% of ads weren’t viewable.

So, three governing bodies of the digital advertising world got together to find a solution. The result is viewability and viewable impressions, measured as vCPM (cost per thousand viewable impressions, a.k.a. viewable cost per thousand impressions).

How is vCPM Calculated?

Glad you asked. Here’s the formula: vCPM = 1000 x Cost / Viewable Impressions

This is all part of a larger initiative to shift from an impressions-served standard to an impressions-viewed standard. In other words, IAB wants everyone to focus on and optimize towards ads viewed versus ads populated.

See this content in the original post

What is measuring impressions?

IAB defines a viewable impression for Display ads as 50% of its pixels are in view for a minimum of 1 second.

Viewability standards for other types of ads:

  • Desktop Video: 50% of the video’s pixels are in view for a minimum of 2 seconds.

  • Large Display: 30% of the ad’s pixels are in view for a minimum of 1 second.

  • In-App Mobile: In-app impressions are generally assumed to be viewable.

While IAB’s definition of a viewable impression is fairly clear, many publishers and vendors have their own ways to determine if and how these criteria are met. Of course, advertisers aren’t too keen on this.

There are many reasons why an impression isn’t considered viewable, such as a user bouncing before the page loads, ad-blocking software, the user doesn’t scroll down far enough, etc.

In 2015, as viewability became the biggest hot button issue, even the walled gardens of Facebook and Google opened up to third parties, in this case, Moat, to prove their viewable impressions were legitimate.

Nowadays, as Nick Lange, Pathlab’s Technical Innovation Lead, says, “Viewability can be calculated by ad servers, DSPs, and third-party vendors.” It’s a balance that helps keep things in check

But as advertisers wielded vCPM and third-party vendors to push publishers to ensure better viewability, fraudsters and bad actors came out of the woodwork.

This leads us to one, perhaps unintuitive fact:

Advertisers shouldn’t necessarily chase 100% viewability.


Let’s get into why.

See this content in the original post

How does viewability work with the marketing funnel?

One way to think about measuring viewability is that it’s a way to determine the general quality of your ad placements. If you have 100% viewability for desktop display, you are almost certainly serving ads in fraudulent websites — or you are being very annoying.

If you think through the tactics 100% viewability requires for desktop ads (mobile is different, check out the Nuance Note above), it quickly becomes apparent why you don’t want it. Unless they’re using lazy loading (explained below), publishers may serve really intrusive in-your-face ads to increase how viewable their inventory is.

Or, in some of the worst cases, fraud like ad stacking makes an ad “viewable” without the user ever seeing it. When you optimize for 100% viewability, you encourage these types of ad placements, so you want to find a goldilocks zone where you’re getting good viewability while avoiding fraud.

And it just so happens this Goldilocks zone aligns well with mid to bottom-funnel campaigns.

Nick explains, “In general, if you’re running an engagement or conversion campaign, you’ll want to hit about 70% viewability. For awareness campaigns, we usually aim a little (sometimes a lot) higher because the goal is to just get the message out.”

He goes on to say, “The overall goal is ROAS (Return on Ad Spend). In general, you can get better CPAs (Cost per Acquisition) below the fold, which means you’re sacrificing viewability to focus on more conversions. Which is great if a good CPA is your goal.”

Why do I get better CPAs below the fold?

The reason you can generally get better CPAs below the fold is that the user is more invested in the content they are consuming. They’re paying attention to what’s on the page and are less likely to bounce. Therefore, they are more likely to ingest your ad’s message and react.

So, as usual, our advice for people looking for recommendations on metrics is: First, think of your campaign objective.

This is a nuance-heavy business and running campaigns is both an art based on human intuition and a science-based on hard numbers and experimentation. Firm recommendations one way or the other will almost always have caveats.

What is lazy loading?

See this content in the original post

Lazy loading (a.k.a. smart loading) is where publishers wait to load the ad until the user scrolls to where it’s in view. So, when you initially land on a page, only the ads you see above the fold load. As you scroll down, more ads will populate as they come into view, rather than all of them loaded at the same time.

This is pretty great for viewability because every ad served on that page is visible to the user (i.e., a viewable impression).

From an advertiser’s perspective, lazy loading combines the strengths of below-the-fold CPA rates with higher viewability. From a publisher’s perspective, lazy loading speeds up load time, can therefore help with SEO, and they can offer (and charge for) 100% viewable inventory.

Increasing viewability is one of the bright spots in the digital advertising industry where publishers and advertisers alike can unite to provide a better user experience while they both benefit monetarily.

See this content in the original post

Viewability benchmarks:

Nick provided a few takeaways Pathlabs keeps in mind when considering viewability.

  1. A 70% viewability rate for mid to low-funnel campaigns is a decent benchmark.

  2. You don’t want 100% viewability on desktop (in-app mobile is a different story, check out the Nuance Note above).

  3. 100% vCPM is going to be expensive and won’t necessarily drive CPA because it prioritizes above-the-fold placements.

Keep these benchmarks in mind as you leverage viewability in your campaigns. While doing so, consider the history of viewability and how it’s measured to truly understand its place in modern campaigns.