You put tons of time into creating your product, experimenting with acquisition channels, and honing your messaging.
Yet here I am, about to tell you that consumers are often swayed by such subtle nudges as the order in which you present your products, or the “serial position effect.”
The serial position effect is the tendency of a person to remember the first and last items in a series best and the middle items worst. It has implications for memory, preference, behavior, and, of course, for designing and optimizing your website.
How the serial position effect influences memory
The serial position effect is made of two parts:
Primacy and recency explained
Primacy: Things that happen first are typically the most important because they influence what comes next.
The theory behind the primacy effect is that considering an item by itself requires a comparatively small amount of processing effort. So, when you process the ninth item on a list, you’re also processing the previous eight.
But you process the first one by itself. This results in greater cognitive fluency and, therefore, greater recall.
Recency: Things that just happened are relevant because they are the most accurate representation of “now.”
The theory behind the recency effect is that items at the end of a sequence are easier to remember because our working memory—the part of our short-term memory that processes conscious and immediate perceptual information—preserves them.
Our working memory holds only ephemeral information, acting as a buffer for new information while it processes it into other, longer-term memory systems.
All that stuff in the middle? That’s likely to be forgotten. Even if people read everything, the stuff in the middle is most likely to be forgotten.
The effect primacy and recency have on recall is powerful and well studied. It’s not a new concept. Hermann Ebbinghaus (1850–1909) originally coined the term “serial position effect” after conducting a number of memory studies on himself.
Then in this study, from 1962, researchers analyzed free recall of word lists ranging from 10 to 40 words. The study supported the effect. Here’s a chart of the results:
As you can see, the last words on the list were much more likely to be recalled. Similarly, as the lists got longer, it became increasingly unlikely that subjects remembered the middle words (as well as the first words).
If list order matters, what’s optimal?
One other finding with the serial position effect is that, if there is a “distracter task” between the information study phase (the list presentation) and test phase (when respondents are asked to recall), the recency effect fades.
However, in the study, primacy effect was still present after a 30-second interference task. From the study, researchers concluded that the capacity of human short-term memory is likely three to four chunks of information at one time.
Be aware of the serial position effect when presenting visitors with a list of any kind (e.g., links, sales pitch, feature list, client list, navigation, etc.)
Baymard gives some advice regarding lists on your website, all good ideas to consider:
- Put the least important items in the middle of the list and the most important first or last. As a commenter on the article said, “I use this effect to add at the end of the list (so bottom of the list) the main attractive item (ex. North European who prefer the South of France) because I know that the users will go to the last item and I know that they will read the first too.”
- If the prospect makes a decision greater than 30 seconds after exposure, place the most important item first. If the decision is made right away, place the most important item last. (This has to do with the time-of-exposure part of the serial position effect.) The example they gave: On a sales page, try putting the main benefit first on the list, and list persuasive extras like “free shipping” and “works with iPhone” last. This way, if the subject leaves the page, they’re more likely to remember the main benefit of your product.
- When the user doesn’t set the pace of the presented items (e.g., as in video and audio), present the most important items last (and first, I might add—especially as they’re probably not going to make it all the way through).
Primacy and anchoring effects
Often, this is the most prominent or memorable option. Serial position effect can aid in this process, especially regarding pricing.
We published a study that looked at two ways of presenting product information:
- Product primacy;
- Pricing primacy.
With product primacy, the subject sees the product first, then the price; pricing primacy is opposite—price first (e.g., you walk up to a rack of $40 shirts at a department store, then see the shirts). The research found:
- When consumers see a price before the product, they evaluate the product’s worth more critically.
- When they see the product first, they evaluate the product solely on that criteria.
Essentially, what the consumer sees first anchors how they judge the entire experience. How can you use this to your advantage? Perhaps with SaaS pricing pages, you can set the highest price first:
Or, similarly, you could test the order of your products on a category page:
You might notice this price anchoring when at a nice restaurant. Restaurant consultants often suggest featuring a super-expensive wine or menu item, so everything else looks reasonable in comparison.
Know, however, that prototypicality is a balancing force. Users are used to seeing an ascending pricing order (lowest first). And if they’re comparison shopping (they are), seeing the highest price first could be a source of friction.
There are always examples that contradict best practices. What works for someone else may not work for you. It’s always contextual, right?
The point is, order can make a big difference, and it’s not a difficult test to set up.
Order affects your preference
The researchers studied “recommender systems”—systems that help consumers choose which product to buy.
Subjects were shown randomized variations of tent order. All the tents had different features (e.g., waterproofing, closures, etc.), yet consumers chose the first tent over the rest by a factor of 2.5X, no matter what the first tent was.
If you’ve ever sent out an email newsletter, you know the order matters. People click the first (and last) links more.
With ours, for example, we feature content from the blog that was published the last week. We consistently find that the most traffic goes to the first one we post.
Other newsletters, like Instapaper, organize by popularity as well:
According to Roger Dooley, this effect has clear implications for website design:
Well, for one, you could put the product you’d most like to sell in front of the others. Perhaps it’s your most profitable product, or the one in which you hold the most inventory.
From a more customer-oriented standpoint, I’d recommend putting your most attractive product up front – the one which offers the best combination of value and performance, for example. This should maximize the chance of an order actually being placed, and should also be the most likely to create a good customer experience (and repeat orders).
Often, the newest products are given precedence, which makes sense because you want a lot of eyes on new products for market validation.
But when designing product pages, try testing different default orders. Try emphasizing the most attractive products. An example? Chubbies places its two most popular products first (and a new product third):
People are simply more likely to click on the first few links in a series, so optimize with that in mind.
Remember this when you’re designing menus, too. We’ve consistently seen that on menus with too many options (i.e. paradox of choice), the middle ones get ignored. Limit your options, or place the important ones at the top or bottom.
What does all this mean For marketing and optimization?
- Start strong. Make sure your value proposition is clear and above the fold (primacy).
- End strong. Repeat the call to action throughout and at the bottom of the page (recency).
In addition, there are a few applications to pricing and other site aspects.
Optimize for the first impression
You know the importance of your value proposition. As Peep once wrote, “value proposition is the #1 thing that determines whether people will bother reading more about your product or hit the back button.”
It should make a strong first impression and anchor the experience for your web visitors. Done well, it should be remembered after browsing your site (and, hopefully, even if they leave).
A shining example of clarity is from Stripe:
What if you’re writing a long-form sales page? Don’t bury the lede. If you want something to stand out in your copy, use it at the beginning. Repeat the message several times, as well.
Our agency site starts strong, and the subhead reinforces that message:
Hard to forget what we do.
End strong: emphasize the core message
The recency effect tells us that our brains also remember things better if they occur at the end of a sequence. Therefore, in sales copy, repeat your most important message toward the end.
Recency also suggests the importance of the last customer touchpoint. I was thrilled when Chubbies sent me well-designed, funny packaging (with some free gifts and a letter). This is how promoters are made.
Conversion research implications and mitigating bias
Anything that requires users to respond to a questionnaire longer than a few questions may suffer from bias due to the serial position effect.
As Jeff Sauro, founder of MeasuringU, writes:
In usability testing, this [Serial Position Effect] is most commonly seen with tasks and product order.
Users typically perform worse on their initial tasks (as they get accustomed to the testing situation and being observed). When testing multiple interfaces or products, the most recently used product may have more salience in the user’s mind when they are asked to select their preference.
The best way to minimize primacy and recency effects is to alternate the presentation order of the tasks and products.
It’s similar with customer surveys. As users begin to “learn” a survey, they tend to lead with answers that they believe will reach a desired outcome. Essentially, they’re trying to guess at the goal of the survey and answer accordingly.
A similar bias to be aware of while designing longer surveys is the bias of central tendency. Fatigued respondents default to the mean score on any kind of Likert or Semantic Differential–type scale.
To mitigate the serial position effect and error of central tendency, just mix up the questions for each respondent. Most survey tools have this capability.
Primacy and recency effects are strong influencers of human behavior. But they’re not clear-cut, and there’s no silver bullet–way to use them. The basic premise stands though:
We remember (and usually prefer things) presented first, as well as most recently. Things in the middle tend to be forgotten or lost in the noise.
Therefore, place things that deserve emphasis first or last on your site. Don’t assume that this will give you a lift, of course—test it. But our memories favor the first and last items in a sequence. Optimize with that knowledge in mind.