How to Protect Against Broken Ads

Jane O'Hara
cardimage

Has a website or app ever welcomed you with a pop-up ad you can’t click out of — or a banner ad that has nothing to do with the page content or your interests?

Whether they’re the wrong size, context, placement, and so on, “broken ads” can damage your user experience, credibility, and advertising revenue. As a publisher, knowing how to prevent them will protect your long-term monetization efforts.

This article will examine the causes and effects of broken ads, as well as offer best practices to protect your sites/apps and your bottom line.

What are “broken ads”?

Broken ads are the cringe-worthy promotions every publisher dreads — the ones that negate all the hard work you did to make your site and app look beautiful.

“Broken ads” here don’t refer to purposefully-obnoxious “bad ad” units — like auto-play, animation, pop-ups, and pop-unders. Most ad networks/exchanges are transparent about the types of ads they show — so a publisher shouldn’t have such ad experiences unless they agree to it.

Instead, a broken ad is a situation where you expect a standard ad to appear (or a standard ad experience to happen) but something goes amiss.

And while you likely know a broken ad when you see one, let’s explore some examples.

Restricted video players

Video ads are lucrative, but if the videos are restricted by your domain or the browser, users will get an unpleasant ad experience.

example of restricted video ad
example of restricted video ad
Ads with broken code

Users of this news app can expect an ad to appear above the top headlines — but not the code for one. This creates a confusing user experience and distracts from the news content.

example of mobile ad with broken code
example of mobile ad with broken code
Dynamic insertion fails

If your ad platform inserts user-specific content — like the user’s location — dynamically into the ad, make sure the null default isn’t obvious. For instance, this Nextdoor sponsored ad for Ring uses “” instead of the user’s specific neighborhood name.

ring ad with failed dynamic insertion
ring ad with failed dynamic insertion
Incorrectly sized ads

Ads that aren’t properly sized for their placements don’t help your users, advertisers, or brand reputation.

It’s unlikely users would want to click on the oversized Xfinity ad below — and equally unlikely that Xfinity enjoys their brand looking unprofessional.

example of incorrectly sized digital ad
example of incorrectly sized digital ad

Similarly, ads that are too small are also unsuccessful. It’s unlikely users can read much of the Audible ad below, much less see the CTA to download the app.

too-small ad for audible
too-small ad for audible
Inappropriate contextual targeting

Ideally, your targeting tactics pair ads according to site content and context, such as an ad for running shoes on an article about a local marathon.

Unfortunately, contextual targeting can go awry without human oversight, as seen in the example below — which matches the content correctly, but lacks the right context.

ad with inappropriate contextual targeting
ad with inappropriate contextual targeting

Correctly nailing content/context is not always easy, especially if you rely on programmatic ads, but for custom ad platforms you can accomplish this through negative targeting features, whereby, in this example, a cruise ship could blacklist its ads from articles containing words like “crippled” and “sank”.

Ad code overload

Forbes may offer good editorial content, but readers are served a tsunami of ads, which can lead to banner blindness and frustrated users.

For instance, those reading the CCPA article below may be too distracted by the number of ads to engage with the actual content.

forbes article with too much ad code
forbes article with too much ad code

It’s of course possible — if not likely — that Forbes is well aware of each and every ad code slot, and this poor ad experience is purposeful. But it may be that the person/team charged with implementing these ad placements isn’t communicating with the site’s designers.

Either way, if you’re striving for a clean ad experience and your page looks like this, you’ll want to troubleshoot where the ads are coming from.

Badly-blocked ads

It’s likely that 25%–40% of your users have an ad blocker. In some cases this will hide the entire ad slot, and the page will look normal. Other times the ad is removed, but the obnoxious user experience remains.

For instance, this ad on Politico.com was blocked, but the big grey box that appears instead is equally obtrusive for readers.

bad ads
bad ads
Malvertising

Unfortunately, broken ads can also be the result of malware — in which malicious code is added to advertising networks — leading to “malvertising” that disrupts what would have been a positive user experience.

These are typically pop-up ads your users can’t click out of, leaving them no choice but to leave your site/app.

Malvertising has appeared on many major sites, including The New York Times and The Atlantic.

How can I prevent broken ads?

Luckily, broken ads can often be fixed — and a commitment to ad testing upfront can lead to higher CPMs (cost-per-thousand impressions) and CTRs (click-through rates) over time.

What to test

Like any of your content, ads require UI testing to be effective.

Improving your ad user experience begins by troubleshooting and debugging — so you’ll need to test all of the permutations that could impact your ad delivery.

These can include, but are not limited to:

Permutations Considerations
Browser/OS With 63K+ browser-platform-device combinations, start with your site/app analytics, which is likely highest for Chrome
Screen/browser/phone size Your ad responsiveness based on mobile device type, etc.
Various screen resolutions Your ads’ appearance based on screen resolution, which can vary by platform, country, and year
Internet connectivity (i.e. 3G, 4G, 5G) Your users’ broadband and mobile speeds, which also vary by country
Individual ad partners The exchanges, data providers, etc. in your delivery
Ad blockers Including different types by browser
VPNs Including usage by region/country
Testing nulls What happens when a user agent isn’t sent (i.e., the Ring ad on Nextdoor)
Targeting features Does frequency capping actually block after X number of impressions? Or is it blocking when it shouldn’t?
SSP/exchange-specific settings Including ad size, format, white/blacklisted domains, and ad categories (i.e., NSFW, firearms, marijuana, “install now”)

1. Browser/OS - With 63K+ browser-platform-device combinations, start with your site/app analytics, which is likely highest for Chrome.

2. Screen/browser/phone size - Your ad responsiveness based on mobile device type, etc.

3. Various screen resolutions - Your ads’ appearance based on screen resolution, which can vary by platform, country, and year.

4. Internet connectivity (i.e. 3G, 4G, 5G) - Your users’ broadband and mobile speeds, which also vary by country.

5. Individual ad partners - The exchanges, data providers, etc. in your delivery.

6. Ad blockers - Including different types by browser.

7. VPNs - Including usage by region/country.

8. Testing nulls - What happens when a user agent isn’t sent (i.e., the Ring ad on Nextdoor).

9. Targeting features - Does frequency capping actually block after X number of impressions? Or is it blocking when it shouldn’t?

10. SSP/exchange-specific settings - Including ad size, format, white/blacklisted domains, and ad categories (i.e., NSFW, firearms, marijuana, “install now”).

Whether your permutations happen programmatically (like in a request to an RTB partner) or they happen in a console, you’ll want to ensure they’re correct before you traffic your ad supply to your ad inventory.

Testing programmatic ads

If you’re serving purely programmatic ads, you’re relying on the SSP (supply-side platform) to detect any bugs.

There are, however, quality control software options available that can serve as a gatekeeper between your ad server and an SSP, such as DoubleVerify and Ad-Juster.

Testing custom/direct-sold ads

Direct-sold ads offer the benefit of human intervention needed to catch many of the broken ads we’ve mentioned, using exploratory testing and other methods that provide solid data.

For instance, a curious, empathetic human ad tester can regulate the number of slots to prevent ad overload (provided it’s unintentional) and add the correct negative targeting features to avoid inappropriate contextual targeting.

Human testers can also help you identify the degree of ad blocking on your site/app, by analyzing your traffic, using Google Tag Manager, or building a custom solution.

Admittedly, controlling your ad experience is easier for Utility Publishers that offer server-side native ads on their own platforms, but all publishers can benefit from ad testing.

What’s the cost of not fixing broken ads?

Given the time and intention you put into creating a great user experience, you’ll want to protect that investment — and your future ad revenue — by ensuring an equally excellent ad user experience.

There are multiple costs associated with broken ads:

Installation of ad blockers

Users know they have the option to block ads — so the less obtrusive and more relevant your ads are, the more likely they’ll be to allow and view them, protecting both your CPMs and CTRs for your advertisers.

Case in point: Remember that Forbes article we shared earlier? Here it is again with an ad blocker.

forbes article with ad blocking
forbes article with ad blocking

As a user, I appreciate seeing only the internal promotional video in the top right corner, but I’ve hurt Forbes’ ability to meet the needs and expectations of the advertisers who make these free reads possible.

User attrition

Because broken ads affect your overall user experience, they may increase your site/app’s attrition rates and cause customer churn.

  • Short attrition rate/lower churn: Shorter/fewer page sessions
  • Longer attrition rate/higher churn: User never returns

Whatever the focus of your site/app content, your users have choices — and preventing broken ads and terrible ad experiences can prevent them from checking out your competitors.

Lost revenue

Broken ad experiences can also lead to lost revenue — not just from the increased use of ad blockers and low click-through-rates — but also from broken ad tags. If a user matching pixel doesn’t fire, that impression could be less enticing to programmatic advertisers. And if an impression pixel breaks, then you have no proof the ad was shown, and you can’t charge the advertiser for that placement.

Malvertising is costly too; ZDNet reported that publishers lost $120M in 2018 to it.

So, while the investment to test ads may seem high, consider the ‘what ifs’ if your ads break. Damage control costs will typically exceed that of preventative measures.

Negative publicity

Broken ads can also lead to bad press, a long-term hit to your brand’s credibility.

For example, trusted news organizations like The New York Times and BBC don’t want to lose their users’ trust by appearing in negative headlines like this one from a 2016 malware attack.

cnet 2016 malvertising headline
cnet 2016 malvertising headline

Additionally, if you have an app, these bad ad experiences could lead to lower app store reviews, making it harder to get new users.

Final thoughts

Preventing broken ads may seem arduous — but controlling your ad experiences allows you to retain your brand’s credibility and maintain high CPMs.

Publishers that fully pre-test and troubleshoot ad delivery will thus offer superior user experiences and enjoy higher profits.

Share your experiences

We’d love to hear about your broken ad experiences and testing methods. Which methods and tools have you found most effective? Click the link and let us know!

Join the Ad.Product community

Sign up for our monthly newsletter and to be notified of member-exclusive events and opportunities.

Ad.Product is the first community for product managers, engineers, and others to discover and discuss how to build innovative, user-first ad platforms.

Jane O'Hara

Recommended Articles