You have twelve features on your pricing page. Your users remember three of them. The question that determines your conversion rate is not which features are best, but which three they remember. And the answer, supported by over seven decades of memory research, is almost always the same: they remember the first feature, the last feature, and whichever feature was most emotionally surprising. Everything in the middle blurs into forgettable noise.
This is the serial position effect, one of the oldest and most robust findings in cognitive psychology. It governs how humans encode, store, and retrieve information from ordered lists. And it has profound implications for how feature lists, comparison tables, and benefit hierarchies should be designed.
Primacy, Recency, and the Forgotten Middle
The serial position effect, first described by Hermann Ebbinghaus in the 1880s, consists of two complementary phenomena. The primacy effect describes the tendency to remember items presented first in a sequence. The recency effect describes the tendency to remember items presented last. Items in the middle of the sequence, the vast majority of any list, are recalled significantly less frequently.
The primacy effect occurs because early items receive more rehearsal in working memory. When you read the first item in a list, you have nothing else to process. You give it full attention and it transfers to long-term memory more effectively. As subsequent items arrive, they compete for processing resources, and rehearsal of each individual item decreases.
The recency effect occurs because the last items in a sequence are still in working memory when recall is attempted. They have not yet been displaced by subsequent information. This is why you can often remember the last thing someone said even if you were barely paying attention.
The practical result is a U-shaped recall curve: high recall for the first few items, low recall for the middle items, and high recall for the last few items. This curve appears in virtually every study of sequential information processing, from word lists to product features to meeting agenda items.
The Hidden Cost of Feature List Order
Most feature lists are organized by one of two principles: technical hierarchy (most complex features first) or assumed importance (what the product team thinks matters most). Neither approach is optimized for how users actually process and remember the information.
When the most differentiating feature is buried in position six of a twelve-item list, it is in the recall dead zone. Users may read it, but they are unlikely to remember it when they make their purchase decision. And purchase decisions are rarely made while looking at the feature list. They are made later, while discussing with a colleague, comparing in their head with a competitor, or simply reflecting on whether the product seems right.
A/B testing data on feature list ordering reveals the magnitude of this effect. Reorganizing a twelve-feature list to place the two most differentiating features in positions one and twelve, without changing any feature descriptions, improved conversion rates by 12 to 18 percent in multiple tests. The features did not change. The product did not change. Only the order changed, and conversion shifted meaningfully.
The Strategic Feature Hierarchy
Applying the serial position effect to feature presentation requires a strategic approach to ordering:
Position 1: Your strongest differentiator. The first feature should be the one that most clearly separates you from alternatives. This is the feature that receives the most processing attention and has the highest probability of being encoded into long-term memory. It sets the frame for how all subsequent features are interpreted.
Positions 2-3: Supporting differentiators. These positions still benefit from the primacy effect, though with diminishing strength. Place features that reinforce the narrative established by the first feature.
Middle positions: Table-stakes features. Place features that users expect but that do not differentiate your product. These features need to be present to avoid disqualification, but their specific recall is not critical to the purchase decision.
Second-to-last position: An emotional or aspirational feature. This position benefits from emerging recency effect. Place a feature that creates an emotional response or connects to the user's aspirational goals.
Last position: Your value anchor. The final feature should reinforce the value proposition in a way that makes the price feel justified. This is the feature users will have in working memory when they look at the price and make their conversion decision.
Breaking the Curve: The Von Restorff Effect
There is one exception to the serial position curve: items that are distinctly different from their neighbors are remembered regardless of position. This is the Von Restorff effect, or isolation effect. A feature that is visually, conceptually, or emotionally different from surrounding features will be recalled even from the middle of the list.
This creates an opportunity to rescue important features from the middle-of-list oblivion. Visual differentiation such as different formatting, icons, highlight colors, or "new" badges can create the distinctiveness needed to trigger the Von Restorff effect. A feature marked with a "Most Popular" tag or presented with a unique visual treatment breaks the monotony of the list and captures attention regardless of its serial position.
However, use this technique sparingly. If every feature has a badge or unique treatment, the distinctiveness disappears and the Von Restorff effect collapses. The technique works precisely because most items are uniform and one stands out.
Applying to Comparison Tables
Comparison tables present a special challenge because they combine serial position effects with visual scanning patterns. Users do not read comparison tables linearly. They scan vertically down the column of the plan they are most interested in, checking for the presence of specific features. This means the serial position effect applies to the vertical feature list, but the decision is influenced by the scanning pattern.
Eye-tracking studies on pricing comparison tables reveal that the first three rows and the last two rows receive significantly more fixations than middle rows. Users also spend disproportionate time on rows where the check marks differ between plans, because differences create the contrast needed for decision-making.
The optimal comparison table design places differentiating features in the first three rows, table-stakes features in the middle, and high-value differentiators in the last two rows. This aligns the information architecture with the natural processing pattern of the human memory system.
The Memory-First Approach to Feature Communication
The serial position effect challenges the common assumption that feature lists should be organized by logical category or technical hierarchy. Those organizations serve the product team's mental model but work against the user's memory system.
A memory-first approach asks different questions. Not "What is the logical order of our features?" but "Which features need to be remembered?" Not "How should we categorize these features?" but "Where in the list will these features be recalled from?" This shift from organizational logic to cognitive reality changes feature presentation from a documentation exercise to a conversion tool.
The features your users remember are the features they buy. The serial position effect tells you exactly where to put the features that matter most. First and last. Everything else is supporting material.