Bleed-over: Response to ads depends on context
A clever series of experiments is reported in the Journal of Consumer Research dealing with how people respond to advertising in context. Researchers Hao Shen, Yuwei Jiang, and Rashmi Adaval had people read pages from a mock magazine that contained articles (movie reviews for a film festival) and an ad (for a watch). Participants were randomly assigned to four groups based on whether the movie reviews were presented in an easy-to-read or hard-to-read font and whether the watch was explicitly associated with the articles (as a sponsor of the festival). The watch ad was always presented in an easy-to-read font.
This seemingly innocent manipulation of the stories (they were identical in content) led to some interesting results. As summarized in a JCR press release:
The researchers found that “as the difficulty of reading the movie review increased, participants found it easier to process the advertisement and evaluated the watch more favorably. However, when the watch was listed as a sponsor of the film festival, the negative feelings elicited by the difficult-to-read movie review apparently spilled over to the watch. That is, as difficulty in reading the movie review increased, evaluations of the watch became more negative.”
There are two interesting findings here, and it may be useful to disentangle them. The first is the fact that making something a little harder to process can make people like it less, regardless of what it says. This is an example of a psychological effect called processing fluency, and it highlights some of the intricacies underlying human emotional processing. The simple fact is: thinking (aka cognitive processing) is hard. Unless we are driven by a motivation or goal that overrides our natural bias in favor of easy cognitive processing (like curiosity, or threat, or satisfying an immediate need or want) things that are just hard to process are going to get tagged with implicit and possibly explicit negative valence (disliking).
A classic example of the processing fluency literature is the wonderfully titled article “Mind at Ease Puts a Smile on the Face: Psychophysiological Evidence that Processing Facilitation Elicits Positive Affect” by Winkielman and Cacioppo (2001). The authors used electromyography (EMG) to measure micro-muscle movements in the face to differentiate responses to images that had been more or less “blurred” by adding some “random noise” in PhotoShop. They found that, indeed, easier-to-process pictures elicited more positive emotions than harder-to-process pictures. Among other implications, they noted:
In many situations, perceivers are required to focus primarily on the semantic content of a communication. The present work raises the possibility that the perceiver’s internal processing dynamics may be a source of affective reactions that have nothing to do with the relevant content. If these reactions are misattributed to the features of the target, they may constitute a potential source of judgmental bias. (emphasis added)
This is exactly what Shen et al. found in their subjects’ responses to the movie festival reviews. People attributed the negative valence they experienced due to processing fluency to the content of the reviews. This result should give pause to anyone in the business of graphic design and presentation, whether for editorial or marketing content.
But the article’s second big finding is even more scary. Processing fluency not only impacted emotional responses to the content, but also bled over into emotional responses to the adjacent ad. So the authors ended up observing a two-step misattribution, first from font to editorial content, then from editorial content to associated marketing content.

These findings raise some interesting questions for the currently popular practice of “native advertising.”
I think the most interesting part of this finding is the fact that the emotional “bleed-over” only occurred when the watch was explicitly associated with the movie festival. It is as if the brain is constantly setting boundaries for what it is going to treat as a single “experience” – if the watch is perceived as part of the movie festival reviewing “experience”, emotional responses bleed over from one to the other, if it is not, the emotional flow stops at the boundary.
The authors, being good academics, are circumspect in identifying implications of their study, noting that “if the magazine content is likely to be difficult to read or understand, it might be better to show advertisements that are easy to read and for unrelated products on adjacent or following pages.” But we, being neither good nor academic, can speculate much more widely, and note many other circumstances in which adjacency of commercial message and media content might be informed by this research:
- Ads and content on web pages (search, entertainment, social, product-related, informational, etc.)
- TV ads in pods during TV shows
- Radio ads embedded in radio shows
- Ads next to news stories in newspapers
- Product placements in movies and TV programs
- Products next to other products on a store shelf
And of course purveyors of content in all these media and contexts have to worry about this effect from the other side of the equation. Can an irritating or hard-to-process advertisement bleed-over its negativity into my programming / article / content / product? Looks like the answer could be yes.
To me, a big take-away of all this is to reinforce the foolishness of proclaiming that the ultimate in-market effectiveness of an ad can be determined by measuring reactions to the ad alone. It’s great to know that an ad provides a great ad-viewing experience, but embed it in a larger TV-viewing experience, or an internet search experience, and its effectiveness can be blunted, or even reversed.
It is content plus context, not content alone, that determines messaging success or failure, and as psychology and neuroscience reveal more about how our brains interpret and interact with the world, we come to appreciate more the challenges of fully understanding what makes persuasive messaging succeed or fail.
The Shen et al. article will be was published in the February 2010 edition but in the meantime can be accessed online here.
Shen, H., Jiang, Y., & Adaval, R. (2009). Contrast and Assimilation Effects of Processing Fluency Journal of Consumer Research DOI: 10.1086/612425
Winkielman P, & Cacioppo JT (2001). Mind at ease puts a smile on the face: psychophysiological evidence that processing facilitation elicits positive affect. Journal of Personality and Social Psychology, 81 (6), 989-1000 PMID: 11761320
First posted in an earlier blog of mine, September 17, 2009. Slightly modified.
About the Author: Steve Genco
