The term ‘viewable impression’ has been around pretty much since the birth of programmatic in the mid-1990s when the first ad servers were developed. Despite being fundamental to our modern understanding of digital advertising, it’s safe to say that the importance of viewability has been largely neglected in the past, mostly due to the lack of reliable measurement methods.
Nowadays, however, the programmatic landscape paints a different picture. With the advertising ecosystem being in a more mature state and competition skyrocketing, advertisers are looking to maximize ROI on their budgets now more than ever. As a consequence, viewability has become a major performance lever, especially for publishers relying heavily on programmatic as their main source of revenue.
While that makes perfect sense on the buy-side, things don’t look so simple on the other end of the line. For publishers and vendors, day-to-day operations have become quite the balancing act. In fact, trying to meet advertiser expectations and at the same time maintain healthy profits presents a number of problems related to yield management tactics.
With the way in-view rates are measured, publishers are often left with a difficult choice - to either restrict volumes in favor of viewability or go with an aggressive scaling approach instead. That said, different technologies emerged overtime to make finding that fragile balance easier. So, today we’ll be looking at one of the most popular methods used to increase impression volumes - ad refreshing, and its relation to viewability.
How is viewability measured? A quick recap
Viewability in general is pretty self-explanatory, so we’re not going to go into its definition, but there are several details on how it’s used and measured that are often overlooked. Let’s have a quick reminder.
The two most important things to remember about viewability are that it stands for a percentage and its value is predictive. This is key to understanding how to optimize for it and how it can be affected by other features co-existing in the same environment.
First of all, increasing the number of viewable impressions doesn’t increase viewability by itself. It only matters how many out of all delivered ads were present on a user’s screen and that makes up the in-view rate (a.k.a. viewability).
It’s also worth mentioning that the viewable impressions generated for each session are only recorded after the real-time auction has already taken place. This means that programmatic platforms are simply reading averages overall recorded sessions, in order to provide predictive data on how certain inventory is likely to perform to buyers.
In summary, viewability doesn’t directly scale up with impression volumes but the more requests you send to your ad server, the quicker you can shift your in-view rates after making system changes. Also, limiting the number of off-screen ads can be just as effective to the overall score as increasing viewable impressions.
Ad refresh and Viewability - Friends or Foes?
In the past, ad refreshing was considered to be ‘the viewability killer’. That was of course true more often than not because of the way the first iteration of reloading ads worked. Today, however, that’s not exactly the case.
Originally, ad refreshing would simply trigger a new request to the server on a fixed timer for all impressions it’s enabled for. While that’s really effective for inflating inventory, the problem is that typically most ads are actually out of view most of the time. This means that by just multiplying all the ad requests, publishers would usually end up increasing out-of-view impressions several times more than their viewable counterparts. This, of course, drives in-view rates into the ground.
Luckily, this is not the only type of ad refreshing available today and it’s actually not used that often anymore. More advanced technologies have emerged in recent years, which allow to get the best of both worlds - keep low-value impressions at bay while dynamically adding new viewable inventory to every page view.
Funnily enough, the former nemesis of viewability is now one of the most popular tools used by publishers to maximize its effects.
Using ad refreshing features to boost viewability
As already mentioned, multiplying inventory across the board either makes no difference to or is likely to diminish viewability. By fine-tuning the way ad refreshing works, however, it can be adjusted to effectively increase in-view rates.
The principle is rather simple. You can target specific user behavior that dictates whether an ad is going to be in view or not, and use that as a trigger for the script. There are multiple ways of doing that based on the specifics of each website, but the basic idea is to only make a new request to the server when the new ad will be in the viewable area.
Such tactics can be extremely effective, because of the fact that viewability is a percentage and it takes into account both the total and the in-view impressions. By shifting the balance towards viewable inventory you can actually deliver better results than would be possible without ad refreshing.
That said, there is more than one way to reload ads, and finding the best fit really depends on the type of website and how users interact with it. Let’s take a look at the most popular ones.
Types of ad refreshing
We previously covered how this type of ad refreshing works, so we’re not going to go into much detail here. It is, however, worth mentioning that there are some instances where this method is still not a bad choice.
If you’re running very few ads per page, your average page depth is rather shallow and you’re required by your partners to refresh at longer intervals, then you might be safe just going with the standard option.
It’s certainly not the optimal way of doing things, but if you don’t have access to more sophisticated ad refresh technologies it’s not going to make that much of a difference.
Despite being less popular than the rest, this method does have its perks. What it does is basically take certain events or actions users take on-site to trigger a new ad call for the selected placements. It could be anything from clicking and scrolling to submitting information or hovering over an element.
This is a method most often used by websites with a lot of clickables and heavy engagement. The main advantage of this ad refreshing type is that it’s very reliable in delivering viewable ads, as user-initiated triggers pretty much guarantee a certain area is in view.
We’ll take a look at some examples later on.
One of the more interesting options we have is ad refreshing based on viewport detection. Basically, this means that the system will analyze what section of a page each user is browsing and use that data to reload ad units.
This method clearly exists to combat viewability issues, but it’s still far from optimal as it falls short when it comes down to accuracy. Having the respective section of a page in view doesn’t necessarily mean that any ads delivered there will be viewable as well. Users often hop on and off different tabs, minimize their browser, or simply leave pages open while doing something else. This leads to similar issues as with the vanilla version of ad refreshing.
In its pure form, viewport-based refreshing is used by very few publishers and ad traffickers.
Smart ad refresh
Last but not least, we have ‘smart’ or ‘dynamic’ ad refreshing. This is actually a collective term that describes a variety of different solutions but generally means that the script is analyzing multiple variables to accurately predict when an ad will be in view.
The most common method is to monitor user activity (i.e. clicking, scrolling, browser usage, etc.), viewable area, and elapsed time. All that paints a pretty precise picture of where and when to show an ad so that it’s in view.
Being the most sophisticated method currently in existence, smart ad refreshing has been adopted by many service providers and has nearly become an industry standard at this point.
Generally speaking, this type of ad reloading can be used in pretty much any situation and deliver great results. Most of the time the question really isn’t whether or not it’s the best tool for the job, but rather if it’s actually necessary for a particular website.
Common use cases
Pretty much all of the mentioned ad refreshing methods have their pros and cons, which means publishers need to make a choice depending on what is likely to work best for their web property. Let’s take a look at several common website archetypes and what makes the most sense for each one.
Article-based websites (news, reviews, blogging)
Recommended types of ad refreshing: Viewport-based, Smart ad refresh
This is the most common type of website right now and it’s characterized by a good page-per-session ration and high scroll depth. As the content is typically distributed over a minimum of 2-5 scrolls, it goes without saying that viewability is a primary concern.
In this situation, ‘smart’ ad refreshing and its less-sophisticated viewport-based alternative bring the right tools to display a new ad only when it’s container has been in view for a certain amount of time. With the former you can also detect when users are actually active on the page, to prevent new ad requests from being generated without engagement, which can damage viewability.
The main thing to consider here is that visitors tend to interact with article-based websites in diverse ways, so it’s very difficult to predict any particular type of behavior. This reflects on the potential results publishers can expect from ad refreshing and the respective impact on viewability. Thus, the better the ad refreshing tool is at identifying what the user is doing, the more it is likely to deliver optimal viewability rates.
Recommended types of ad refreshing: Event-triggered, Smart ad refresh
There are currently web tools out there that do pretty much anything, from interactive maps to graphic editing tools. All of these have their unique features, but it’s safe to say that most of them have something in common - everything fits on the user’s screen, or in other words, there’s no scrolling.
This creates an interesting situation when it comes to ad refreshing and viewability because technically every new ad delivered should still be in view, right? Well, not exactly.
Web apps usually bring super long session durations and users often leave them open in the background, especially those used for professional purposes. This means that not every impression is going to be viewable after all.
On top of that, such websites often are more performance-intensive, so anything that drains processing power will be problematic.
That’s why most publishers turn to use events and clickables as a trigger for the ad refresh function to initiate. It’s a pretty reliable method to detect real page engagement, while also being extremely lightweight.
Using this method, publishers can also control exactly when a new set of ads should be displayed, instead of just using a timer. One such example is pressing the ‘save’ button while using an online photo editor. At that point, the visitor is pretty much done with the app and it’s safe to initiate the ad refresh without impacting usability and CPU levels.
Predictably, ‘smart’ ad refresh scripts are also suitable here as they pretty much have all the functionality of the event-based method built-in. They also provide more flexibility, as they have virtually all the features one could ever need to experiment with. So, that would generally be the safer bet in the long-run, although it might not make that much of a difference in this case.
Recommended types of ad refreshing: Time-based
Download libraries are increasing in popularity and are also quite different from the rest in terms of usability. Typically publishers would see shorter sessions but at the same time a significantly higher number of page views per session.
Some might argue that this makes the whole concept of ad refreshing obsolete here, but that’s not necessarily the case. You can still take advantage of those visits where users tend to linger and with the way these sites are typically set up viewability is likely to get a nudge without much effort as well.
With that in mind, publishers might as well stick with the traditional, time-based ad refresh method as pretty much none of its disadvantages are likely to be an issue here. In theory, any type of ad refreshing should work just fine in this environment, but this one is the easiest to set up.
Technologies that work well with ad refreshing
Last but not least it’s worth mentioning that there are several popular tools online publishers can use to see an improvement in viewability, which work particularly well with ad refreshing.
Lazy loading is one such example, which everyone should be familiar with in the current day and age. While advanced refreshing mechanisms can increase viewability for incremental inventory, they don’t have any impact on the first impressions generated on the page. In a nutshell, lazy loading allows you to ensure pretty much every single ad is in view.
Sticky functionalities can also be quite useful. On their own, such placements don’t have a better viewability score than static placement, but with ad refreshing enabled it’s a different story. The sticky script ensures that the ad unit stays on-screen, while the ad refreshing tool takes care of the rest.
And finally, we have the infinite scroll. This shares the same benefit as lazy loading, by creating more viewable inventory to begin with, but it also has another great feature. As opposed to its predecessor (content pagination), it doesn’t wipe out anything already loaded on the page This gives the ad refreshing script more time to go through its timeout interval and potentially get a few more ticks in when users are browsing back and forth.
In summary, it’s apparent that ad refreshing and viewability are no longer mutually exclusive and there are in fact multiple ways publishers can use one to improve the other. While there is usually more than one way to achieve that, it’s not that complicated once you understand your audience and how they interact with your website.