Sept. 15, 2017 by Sara Jabbari
When it Comes to Fill-Rate, Less is More
What is fill-rate and why is it important?
Whenever there is an ad opportunity and an ad request is made to an ad source (such as an ad network) it can either be returned, i.e. filled or not. The fill-rate is therefore the percentage of ad requests that are successfully filled based on the total number of ad requests made. As impressions are monetized on a CPM basis, it may seem logical for a publisher to want a fill-rate as close as possible to 100%, implying that the publisher is taking full advantage of every revenue opportunity. However, this metric represents only one side of the coin, and maximizing revenue does not always mean serving the highest possible number of ads. In the case of the fill-rate, sometimes less really is more.
Which factors influence fill-rate?
Before we explain why, let’s explore a number of factors, both external and internal, that influence the fill-rate.
- Supply-demand imbalance: This imbalance can be attributed to the number of ad networks running, the specific geo targeting of campaigns, and the floor CPM price being set too high to find an ad to serve.
- Placement: the type, placement and viewability of ads can affect the fill-rate. In-stream video ads are more popular than out-stream units while pre-roll inventory is demanded more than mid- and post-roll inventory.
- Device: As web traffic and digital consumption move to mobile, ads are also more likely to be served on mobile devices than desktop.
- Tech errors: There can be internal technical errors and breakdowns that prevent an ad from being served. This includes latency issues with the website or ad server that can lead to timeouts, esoteric ad formats, using http vs. https etc… These technical issues prevent an ad from being served or can cause it to load so slowly that the user exits the site before the ad can be displayed.
Why shouldn't I want to maximize my fill rate?
Due to so many possible scenarios, many of which are outside of publishers’ control, it is essentially impossible to reach a 100% fill-rate. Still, many publishers attempt to maximize their fill-rates, recognizing that every served impression has an associated price tag, and therefore represents a source of revenue. While a high fill-rate can generate short-term revenue, this metric has much bigger long-term implications that indicate that simply focusing on increasing your fill-rate actually creates a race to the bottom, where you push your ads and end up alienating your audience, affecting your overall performance and revenue stream.
Let’s take a closer look. The amount of revenue a publisher generates is determined by multiplying three components: the number of impressions, the fill-rate and the CPM.
At first glance, it may seem straightforward that the higher the fill-rate is, the greater your revenue will be. However, these components are not exclusive from one another. If a publisher’s fill-rate is very high, it can be serving an overwhelming number of ads, strongly deteriorating user experience. What good is a high fill-rate if your users are unhappy and less likely to revisit your site in the future? As a result of this, the first component of the equation (number of impressions) falls, along with revenue. Not only is this bad for publishers, but the marketing messages of advertisers make less of an impact than they would if they were targeted at the right audience, rather than just the largest one. In fact, a poor user experience can even lead to an increased adoption of ad blocking, meaning ads served on a site are never actually seen by users. As fill-rate does not take into account ad blocking, its value can often be misleading.
Not only does fill-rate affect video views, and thus the potential number of impressions, but most publishing sites that have high fill-rates typically set the cpm floor very low in order to meet high demand. By focusing on just one metric (the fill-rate), they miss out on the optimum combination that would maximize revenue, while providing an experience that encourages users to come back to the site. Moreover, by reducing the number of ads served on your site, the available inventory that is sold to advertisers may be considered as premium and command higher CPMs since it takes into account the quality of user experience.
If we come back to our revenue equation ($ = impressions x fill rate x CPM), we can see that while your initial impulse may be to maximize your fill-rate, this strategy is short-sighted. The long term implications on user experience will play a large role in shaping the value of your inventory to advertisers, and your site to users. So, while some may argue that a low fill-rate points to under-monetization of video content, it is really a question of whether the short-term gain is worth the long-term pain.