Watch this trailer of The Hateful Eight, and then fill in the blanks:
K I _ _
Because of priming, you’re most likely to spell k-i-l-l because of the violent images in the video clip. You’re highly unlikely to fill in the blanks to spell k-i-s-s or k-i-n-d, or anything else really.
Priming is a popular strategy to influence human behavior, and it’s one that has been studied extensively for quite a while.
Table of contents
What Is Priming?
Priming is “an implicit memory effect in which exposure to one stimulus influences the response to another stimulus.”
In other words, priming takes implicit (not totally conscious) memories or associations and uses that to influence the response to something else, either related or not. Your previous ideas about the construct of a camel allow you to finish this sketch before it’s actually finished:
Perhaps the most widely known (and hotly debated) priming study was published by John Bargh in 1996. If you’ve read Malcolm Gladwell’s Blink, you’ll recognize the setup:
Researchers primed participants with a list of words associated with stereotypes of elderly people: Florida, old, lonely, grey, selfishly, careful, sentimental, wise, stubborn, courteous, bingo, retired, wrinkle, rigid, traditional, conservative, knits, dependent, ancient, etc. The control group was given a list of neutral words. What they found was that the participants that were given the elderly-associated words actually walked slower when leaving the experiment than did their control group peers.
Another experiment by Bargh found that participants who had been primed with words related to rudeness were more likely to interrupt the experimenter.
There are several types of priming, a few of which are more specifically related to growth and optimization.
Types of Priming
Here’s a list of the different types of primes (as put together by changingminds.org):
- Conceptual priming – when related ideas are used to prime the response (e.g. ‘hat’ may prime for ‘head’).
- Semantic priming – when the meaning created influences later thoughts. Semantic and conceptual priming are similar and the terms are sometimes used interchangeably.
- Non-associative semantic priming – related concepts but one is less likely to trigger thoughts of the other (e.g. ‘Sun’ and ‘Venus’).
- Perceptual priming – based on the form of the stimulus, for example where a part-picture is completed based on a picture seen earlier (like the camel example above).
- Associative priming – when a linked idea is primed (e.g. ‘bread’ primes the thought of ‘butter’).
- Masked priming – when word or image is presented for a very short time but is not consciously noticed.
- Repetitive priming – repetition of a word or phrase leads to influencing later thoughts.
- Reverse priming – when people realize they’re being primed and overcorrect in the other direction.
There are some highly interesting types of priming, but for the purposes of user experience and optimization, the most applicable types are Perceptual Priming and Conceptional Priming, and perhaps Repetitive Priming in long form copywriting.
More Popular Priming Studies
If you’ve been learning about conversion optimization for long enough (or if you’ve just got great taste in book), you’ve probably read or heard about Thinking Fast and Slow by Daniel Kahneman.
It’s a bible of behavioral economics, showcasing many fascinating studies around the seemingly irrational behavior of humans. Lots of it has to do with priming.
For example, when completing the following words, “w _ _ h” and “s _ _ p” , he found that, after being asked to think of something they’re ashamed of, people are more likely to fill them in as “wash” and “soap.” A similar experiment found that after thinking about stabbing a coworker in the back, participants were more likely to buy soap, disinfectant or detergent than other items like batteries or gum.
There was also an interesting study where researchers primed participants with God concepts, that they were being watched by a higher being. They found that subjects “allocated more money to anonymous strangers when God concepts were implicitly activated than when neutral or no concepts were activated.”
Weird stuff (and controversial – more on that later), but what’s this have to do with making more money online?
What Does Priming Have To Do With Optimization?
Priming, as we’ve seen, has some neat applications to real world scenarios. It’s also got practical use cases when it comes to optimization, notably in improving the user experience by ‘nudging’ users in a desirable direction.
If you use priming correctly, you can “key a user in on a certain way of thinking, or predispose (i.e. priming) a user to make certain choices by the rhetoric you use.”
There are three pretty practical ways you can use priming to increase user experience, and therefore conversions:
- Avoid accidentally priming negative associations
- Prime for prototypicality and expected experience
- Prime for desired actions on the user flow
1. Avoid Accidental Priming
First, page elements can prime user behavior and expectations. Sometimes, you can accidentally prime for negative experiences. NN/g gave the great example of a coupon code on an eCommerce site:
So if I’m shopping on this site, and I see the coupon field, then suddenly I’m thinking about saving money with coupons. It might not have even crossed my mind before that. The presence of that box, however, primes me to leave the checkout flow and search for a promo code. And if I can’t find one, I’ve got FoMO (fear of missing out). Cue cart abandonment.
Another example of this is mentioning ‘trigger words’ that cause a visceral reaction. For example, spam. As mentioned in our article on credibility, if a cleaner came to your house and mentioned they weren’t going to steal anything, what would you immediately think? That they were going to steal something! Mentioning the word spam near an email sign up could have the same effect. Case study from ContentVerve.com:
2. Priming for Expected Experience
Priming studies also suggest we should choose our images and words with diligence. Our visitors make decisions based on their instant first impressions, so choosing elements that accurately reflect the message we want to project is important. NN/g gives the following example:
This is a website for a private school that includes up to grades 8. However, the only images displayed are those of kids that look really young, maybe preschool age. This might make you believe the school is only a preschool, or at best elementary school.
One more point from NN/g is that the look-and-feel of a page can create first impressions about a website and what it’s product/service is like. This is largely based on preconceived notions or stereotypes about how a certain website should look (in other words, prototypicality).
This used to be a much greater disparity between local business websites and national chains, but Squarespace, Wix, and cheap & easy WordPress themes are bridging the gap a bit. Still, there’s a certain ‘look’ to a local business, and you recognize it pretty fast:
3. Priming for Desired Action
Another use case of priming is more applicable in designing for the user experience. If you want users to complete a specific action after the onboarding flow, you can prime for that with some strategic image placement. David Teodorescu gave a really good example of creative priming in a recent article:
These are just a few ideas and examples. In reality, the possibilities are endless. For more practical ideas on priming, check out this article from Usabilla.
One thing to keep in mind, though, is individual subjectivity. I’m sure plenty of people could take an example such as the “elderly” study above, and extrapolate that in ways that wouldn’t work – like in the slippery topic of “color psychology.”
You may hypothesize that blue colors are associated with calm and soothing emotions, and therefore you can prime your visitors to feel that way by using such colors. Probably not going to work, and if it does, there’s absolutely no way to attribute or measure anything, so it’s probably just a story you’re telling yourself.
The Limitations of Priming
Of course priming isn’t a silver bullet. Nothing is. Actually, there’s a lot of debate as to the actual effectiveness of using priming outside of the academic world.
Gary Gutter, a professor of philosophy at the University of Notre Dame, voiced his concerns in a NY Times article:
“The reason is that priming experiments seldom tell us how important priming is in realistic situations. We know that it has striking effects under highly simplified and controlled laboratory conditions, where the subjects are exposed only to the stimuli that the experimenters provide. But it is very difficult to know how significant priming stimuli (thinking about money, large numbers, abstract questions) would be in a real-life, uncontrolled environment, where all sorts of stimuli might be conflicting with one another. Also, there is seldom any reason to think that even a strong priming effect will last very long.”
By the way, that’s not only a problem with priming. It’s an issue known as ecological validity – the extent to which the findings of a research study are able to be generalized to real-life settings. If a study has high ecological validity, it can pretty safely be applied to real world contexts. If it has low ecological validity, it can’t be.
Anyway, Gutter raises some good points, notably because almost all of the famous priming studies have failed to be replicated. For instance, researchers repeated the slow-walker study and found no difference in the rates of walking between goal-primed and unprimed subjects.
However, the criticisms aimed at priming studies seem to address, as the New York Times put it, “how best to elicit and measure intriguing properties of the unconscious and, more broadly, how to ensure that the tenets of science — transparency, skepticism and self-correction — manifest themselves in its conduct.”
In other words, priming is not the exclusive owner of these replicability problems.
As The New York Times said, “This is because, in a variety of fields, subtle differences in protocols between the original study and the replication attempt may cause discrepant findings; even little tweaks in research design could matter a lot.”
Overall, it appears that priming, while not a one-size fits all model, does work. It some cases, it works really powerfully. Sometimes, it can’t be replicated.
Priming isn’t only useful in amusing psychological studies or in cocktail party conversations – it also has powerful uses in both online optimization as we well as qualitative research.
First, in optimization, microcopy, icons, and other elements can be used to prime certain actions. This can be a positive or a negative. For example, a ‘coupon code’ area can prime people to expect a discount (and to go searching for one). If they can’t find one, they’ve suddenly got FoMo.
Or you can prime users to expect a certain level of service with good clean design and practical navigation.