Native Ads, Good or Bad?
If you’ve browsed the web at all in the past 5 years, you’ve likely come in contact thousands upon thousands of native ads. I’ll spare you the explanation of what native ads are — also known as recommended/sponsored content ads — and get right to the part everyone wants to know. What does the data tell us about native ads, and are they good or bad for your website?
It’s kind of a loaded question. Good and bad are a little subjective in this instance because it really depends on what we’re measuring. For the sake of this article, I’ll be spending my time digging into the data and learnings we have from publishers and websites that have run different experiments with native ads within the context of how they’ve affected user experience and monetization.
I’ll highlight what publishers should think about when considering native ads, how we’ve seen them used successfully, and the known risks associated with displaying them on your site to visitors.
How do native ads affect other ads?
Let’s start by talking about the whole reason why you might be considering — or currently using — native ads to begin with. Can they help you earn more digital revenue, and is it worth it (however you want to slice it)?
The answer is loaded with caveats. Sometimes, yes. Sometimes, no. One of the biggest things you really need to understand before we dig into all the caveats is just how native ads affect other ads on the page.
This is something every publisher should really be measuring. It’s really important to understand at a fundamental level that all ads affect and dilute all other ads on the page; this is especially true for native ads.
Native ads are designed to look like the relevant content or additional content, so they will — in some way — affect your user experience metrics (which have a scientific correlation with your overall revenue).
This isn’t good or bad on its own, but must be constantly measured and optimized against to ensure you don’t do long-term damage to your site’s earning potential. For example, that native ad click may have good value, but without the native ads, would visitors be visiting an additional 2-3 pages? Would those pages be totaling more revenue from ad impressions than that single click on a native ad?
This is really the math you have to start with. Our data science team has learned that native advertisements have the potential to steal or dilute the other ads on page in a way that can often be misleading.
For example, native ads may be reducing the CPM’s of your other display ads in a way that makes it seem as though the earnings from the native ads are the highest earning on the page. However, when native adverts are taken away, the other ads on the page will actually earn higher CPM’s. This is why it is CRITICALLY important to understand EPMV (earnings per thousand visitors) and how landing page revenue for any page with native ads is affected by this process (if you don’t understand that, check out this info on landing page revenue and EPMV).
Takeaway: How native ads affect overall session revenue is far more important than how much they earn on their own.
Native ads affect user experience — this is critical to understand
More so than how native ads affect other ads — and potentially several other things we’ll discuss below — is the impact that native ads have on user experience. Native ads almost always have some kind of effect on bounce rate, session duration, and pageviews per visit.
Again, this information on its own is not good or bad. You have to understand how it works on your site. Every site is different, but more importantly, every visitor is different.
Publishers should only be showing native in the proper context (visitors that aren’t affected by native, people from social media sites, etc) and not to everyone (organic search visitors, etc). These are some generalities but it’s part of what we’ve learned. Native ads make more sense for certain types of users than others (social vs organic visitors being a common one).
For example, on one site that we work with, native ads ultimately had a negative impact on pageviews per visit and bounce rate. This is really important to understand because it has also has a negative impact on total session revenue; which is obvious once you understand the correlation between those two things.
I am in a unique position because Ezoic works with thousands and thousands of different kinds of publishers, so we’ve learned that some types of sites see better results with native than others, but what determines better almost always comes down to user experience.
Native ads may negatively/positively affect some sites more than others. That’s why actually testing native ads on your site is really important, and if possible, only delivering them to the users that are more likely to not be negatively affected by them.
An example of a site that might tolerate native ads well is a celebrity news site that generates 50% of traffic (or more) from Facebook — although native advertisements may be falling out of favor with Facebook. Meanwhile, a science-based website discussing astronomy news may not have a user base that is accepting of many native-style ads.
How do we identify users that are negatively affected by native ads?
That’s a great question. It’s actually fairly tough to do and implement without a multivariate platform or some kind of A.I. but here is how you could start…
- Test versions of the site with multiple native placements vs. a version of the site without
- Segment users in Google Analytics by geo-location, UTM source, and seasonality
- Compare apple to apples of UX metrics for each of those segments vs. each version of the site you’ve tested
- See how the native versions of the site impact bounce rates, pageviews per visit, and session duration for different segments and create proxied versions of your site that deliver only to the ones that were not negatively affected.
Takeaway: Know how native advertisements work on your site and how they affect your users. Testing is a great place to start.
Native advertisers want to leverage your brand
Native advertisers are essentially attempting to borrow your visitors by using your brand/content/platform as a leverage point. People think that the content is recommended by your site, so they are more likely to click.
This is as clear of a tradeoff as you can imagine. For many major media brands, it’s a no-brainer, most forms of native ads are too dangerous to their brand to implement. However, as you can see above, native advertising has grown considerably and is more accepted than ever before. Softer versions of native ads are being built to try to bring back brands that are concerned about their image with native ads.
Publishers must be able to truly understand the data on their site as it relates to native ads in order to make these tradeoffs make sense. Native ad providers will always push for more guaranteed prominent placements that dilute other ads, user experience, and your brand.
As we learned above, this is OK for subsets of users and other specific situations, so it’s all about being smart with your implementation. However, there is a lot to consider before blanket implementation is completed. Just about any publisher can probably use them as a tool in their overall monetization strategy, but it all has to do with context. The secret is having control over the ads themselves.
The real danger that publishers face is contracts or misleading implementation of native ads by providers. Many providers attempt to offer big upfront paydays to publishers for prominent placements on their site, then, they implement units that include additional link units and things like auto-play video that dilute other ads and user experiences .
The money up front is tempting, but how much could be potentially lost it that drags avg. pageviews per visit down? Remember the chart from above, those additional page views often represent 100-200% uplifts in session revenue. Are the native ad providers offering 100-200% more digital revenue than you could have made off of existing resources?
Takeaway: Maintain control over how and when you display native ads and apply a data-driven method of implementation.
Testing is critical
If there’s one important lesson to learn from this article it is the variability of how sponsored content ads vary depending on a lot of different data points. Testing these ads on your site and your users is critically important to understanding the impact that native ads have on user experiences and revenue.
If you don’t collect or understand this data you could be dealing with a major risk. Furthermore, its very important to maintain control of these types of ads so that you can ensure that you are leveraging in line with tradeoffs between brand, revenue, and session length. Data and control are the two most important components to the process.
Have questions? Leave them below and I’ll share any information I can.
