Mental Models
Mental models are the assumptions shoppers bring to your site before they read a word of copy. When your UI breaks them, conversion drops — here's how to spot and fix it.
Mental Models
The internal predictions shoppers use to anticipate how a site, product, or checkout will behave before they interact with it.
A mental model is the rough cause-and-effect map a visitor carries in their head — built from every other store they've used — that tells them where the cart icon lives, what "free returns" usually means, and what should happen when they tap a product image. Designers don't install mental models; shoppers arrive with them.
In CRO terms, mental models decide whether your interface feels intuitive or broken. When your filter UI, PDP layout, or checkout sequence matches the model the visitor already holds, friction disappears and they convert on autopilot. When it diverges — a non-standard size selector, a hidden shipping cost, an unfamiliar payment step — cognitive load spikes, hesitation grows, and drop-off follows.
Mental models sit downstream of memory and perception: shoppers compress thousands of past sessions into a handful of expectations they apply to every new store. That's why a first-time visitor to your Shopify apparel site already "knows" where to find size, price, and the bag icon — they're pattern-matching against Zara, ASOS, and the last twenty checkouts they completed.
The trouble starts when your store is novel in ways that don't pay off. A clever mega-menu that hides the search bar, a PDP that puts reviews above the buy box, or a checkout that asks for an account before showing shipping cost — each breaks a learned pattern. The visitor doesn't think "interesting choice"; they think "this site is off" and bounce.
Friction Score = (Unexpected Elements / Total Decision Points) × Severity Weight
Unexpected Elements
Mismatched interactions
Count of UI patterns on the page that diverge from the dominant convention in the category (e.g. unusual filter behaviour, non-standard add-to-cart).
Total Decision Points
Total interactions required
Every step the shopper must take to complete the task — selecting variant, entering address, choosing shipping, etc.
Severity Weight
Impact multiplier
1.0 for cosmetic mismatches, 2.0 for trust-related (price, shipping, returns), 3.0 for blocking (can't proceed).
A beauty store's PDP-to-checkout flow has 8 decision points. Three diverge from category norms: a non-standard shade picker (cosmetic), a shipping cost revealed only after address entry (trust), and a forced account creation (blocking).
Unexpected Elements (weighted): 1×1.0 + 1×2.0 + 1×3.0 = 6.0
Total Decision Points: 8
→ 0.75
A friction score above 0.4 typically correlates with a 15-25% drop in checkout completion versus a category-aligned baseline. This flow is bleeding revenue at every weighted mismatch.
Severity matters more than count. One blocking mismatch (forced account creation, surprise shipping fee at step three) destroys more conversions than five cosmetic ones combined, because it violates a model tied to trust rather than navigation.
Typical conversion impact when a UI element breaks the dominant category mental model
| Surface | Common mismatch | Estimated CVR impact | Fix difficulty |
|---|---|---|---|
| Product listing page | Filters don't update count live | -4% to -8% | Medium |
| Product detail page | Variant selector behaves unusually | -6% to -12% | Low |
| Cart drawer | Hidden subtotal or shipping estimate | -5% to -10% | Low |
| Checkout step 1 | Forced account creation before guest path | -15% to -25% | Low |
| Checkout shipping | Shipping cost revealed only after address | -8% to -14% | Medium |
| Returns page | Return policy hidden in FAQ instead of PDP | -3% to -7% | Low |
The fastest way to audit your own store is to watch five session recordings of new visitors completing a purchase. Every hover-then-pause, every back-button press, every scroll-up to re-check something — that's a mental-model mismatch surfacing in real time. Patterns repeat across users; one or two mismatches usually account for most of your drop-off.
Mental models FAQ
A persona describes who the shopper is (demographics, goals, context). A mental model describes what they expect your interface to do. Two very different personas — a 22-year-old and a 55-year-old shopping for sneakers — often share nearly identical mental models for how a PDP should behave, because both have used hundreds of similar stores.
Mental models are the output of memory and perception working together: perception captures the current screen, memory supplies the pattern library of "how stores like this usually work," and the model is the prediction the brain generates by combining them. Break the prediction and the shopper feels friction even if they can't articulate why.
Rarely worth it. Novelty in checkout, filters, or pricing almost always costs conversion. Novelty in brand storytelling, packaging reveals, or post-purchase emails can win — those are surfaces where surprise is welcome. Keep functional UI conventional; reserve creative risk for emotional surfaces.
Three sources: session recordings (look for hover-pauses and back-presses), exit-intent survey responses ("I couldn't find..."), and a five-second test against three competitor sites in your category. Anywhere your store diverges from the category norm without a clear payoff is a candidate.
Yes, but less than you'd think. Payment methods (iDEAL in NL, Klarna in DE/SE, Cash on Delivery in IT) carry strong regional expectations, and address-form order varies. Most PDP and listing-page conventions are global because the dominant marketplaces — Amazon, Zalando, ASOS — have trained the same patterns everywhere.
The models are surface-specific. On mobile, shoppers expect a sticky add-to-cart, a bottom sheet for variants, and a thumb-reachable cart icon. On desktop they expect a left-rail filter and hover states. A pattern that matches desktop conventions can violate mobile ones — audit each separately.
Jakob's Law is the operational rule: users spend most of their time on other sites, so they expect yours to work the same way. It's the design principle that follows directly from how mental models form — you're not designing in a vacuum, you're designing against a baseline set by every other store in the category.
Yes, and the wins tend to be unusually large and fast-to-significance. Mental-model fixes — exposing shipping cost earlier, adding a guest checkout, restoring a conventional filter — typically produce 5-15% lifts in the affected step, visible within 1-2 weeks on stores doing €1M+ revenue.
They shift, but slowly. New conventions take 3-5 years to become dominant (Apple Pay buttons, swipeable image galleries, sticky CTAs). Trying to ride a convention before it's dominant is risky; adopting one after the top three category leaders have is safe.
Exposing total cost — including shipping and any fees — before the final checkout step. Surprise cost at the last step is the most violated and most expensive mental-model mismatch in e-commerce, and fixing it typically recovers 8-15% of abandoned carts within a month.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.