And recent findings suggest that, in some cases, annoying ads can literally cost publishers more money than they generate, leading to a cost of up to $1.50 CPM for ads that may net the publishers well below that amount.
A group of Microsoft researchers published their findings in last December’s issue of the Journal of Marketing Research. Specifically, they calculated the economic cost associated with presenting annoying ads to readers.
To do this, they first compiled a set of 144 ads, which they ranked in terms of how annoying these ads were to a sample of 141 users. From this, they selected the 10 most-annoying and 10 least-annoying ads.
Next, they designed a field experiment in which an additional set of 1,223 users were asked to read a Web page and to categorize its content into one of four predefined categories. The content on the page could be presented by itself, or it could be flanked by skyscraper ads, which were drawn either from the more-annoying or less-annoying set. Users were paid for each item they categorized (regardless of whether they categorized it correctly or incorrectly), and they could choose when to stop.
Not surprisingly, users who had to deal with the more-annoying ads viewed a lot fewer pages (“impressions”) than those who had to deal with less-annoying ads, or with no ads at all. This provided an estimation of the economic value of dealing with annoying ads.
Through a carefully controlled experiment design, the authors were able to estimate the incremental cost of 1,000 impressions with the more-annoying ads, relative to a page with no ads, at $1.53. Even when comparing more-annoying ads relative to less-annoying ads, the incremental CPM cost was $1.15.
Given that many of the more-annoying ads also tended to be associated with lower-quality ads, which typically fetch well below $1 CPM, publishers would be economically better off not showing annoying ads. That’s because the number of page views they lose from annoyed visitors outweighs the monetary benefit of the ads delivered.
Furthermore, although this particular study focused entirely on aspects of an ad itself that made it annoying, other factors such as page clutter, pop-ups and takeovers can further increase visitor annoyance even if the quality of the ads is higher, leading to even greater “hidden” economic costs.
Even more importantly, these results are based strictly on the immediate economic value of an impression, and do not take into account long-term effects such as lowering the visitor’s perception of the publisher site, or the likelihood that visitors, annoyed by the ads, will not return.
While the details of these results surely depend on the specific ad, the website, the visitor, and many other factors, the message is clear: Maximizing ad revenue by filling more inventory with low-quality, annoying ads can potentially have a negative economic impact even in the short term -- let alone the long-term negative consequences to the publisher’s brand.
Interesting study, thanks Paolo. The article did not offer much detail about what makes an ad "annoying." Can the author shed some light on what common annoyances users noted?
Good story Paulo. One of the big problems is you can block the bad ads on Google Adsense but they will simply change the URL a few days later. Then when you hit the 500 maximum ads you can block, Google will not do anything. You are either stuck with Google's trash ads or spend a ton of time trying to manage the problem.
The other big problem with the bad ads is some will run malware and super cookies behind them. This is what really upsets my members.
Thanks Andrew, glad you found it interesting. The authors of the study cited prior research suggesting that of the many factors that make an ad annoying, animation was the most prominent - especially "fast animation." Unfortunately they did not explain exactly what sort of animation, but they describe a process in which they started with a set of 72 animated ads, and built 72 more by taking the final frame of the animation. They did a preliminary study with 163 subjects and asked them to rank the level of annoyance of ads on a scale from 1 to 5, with 1 being "much less annoying than the average ad in this experiment" and 5 "much more annoying than the average ad in this experiment." They found a very stong correlation between animation and annoyance with the 21 most annoying ads all being animated and the 24 least annoying all being static. They also found that the avarage annoyance level for animated ads was 3.6 vs. 2.4 for static ads. And they also asked subjects to give a brief description of why ads were annoying, and found that variants of "animation" was the most frequent complaint.
If you would like more details, the full reference to the article is: D. G. Goldstein, S. Suri, R. P. McAfee, M. Ekstrand-Abueg, F. Diaz: “The Economic and Cognitive Costs of Annoying Display Advertisements.” Journal of Marketing Research, V.51(6), December 2014.
@Craig - you are spot on about ad blockers. The authors also cited a statistic indicating that there are staggering numbers of ad-blocker downloads. The recent problem with malware that can be spread through ad units distributed through RTB is definitely a concern as well. In all, it seems that everyone agrees that ads can be damaging, but it's important to start quantifying these effects in the hope that at least some publishers will begin to be more strategic about it.
Interesting and factfull, no publisher give a thought on hidden cost which they at times actually lose and they don't realise till they see huge difference in the profit.
But reading the article raised me a question, what if i show interesting and creative ads only? If the ads are not intrusive i don't think so it will affect a lot.
And if a publisher don't run ads they will end up nothing at the end, so the fact of hidden cost is acceptable but running no ad or less will eventually lead to no earning ( this goes for content related to sites)
To add i really feel rather then stopping running ads we better make use of ad verification tools like GeoEdge or any other in the market. I say GeoEdge beacuse i have tried and worked good for me (no promotion but just sharing experience).
Using ad verification can help blocking specific ads which are actually hurting the image of the site plus irritating our users, and this will lead to no loss at the end.
Paolo, another intesting problem has started in the past couple of months. With one of my RTB's I have ads totally disappearing. After doing weeks of checking we are pretty sure the problem is since we use a ad rotator between 3 companies, when the one company's is showing in more than one of the three positions, then only the first location is shown and the other one or two does not. While this might sound a little trival, but when you run millions of pages a month, this starts to ad up quite a bit. Worse the junk ads become more noticeable.
@Avinash - you are correct. The suggestion of the study was not that publishers should nor run ads at all, but simpy that annoying ads have a clear economic cost. For brevity I only mentioned this in passing, but the authors of the study did compare annoying ads to less annoying ads and found that even in that case there is a clear economic cost to the mroe annoying ads.