Anatomy of a product page that converts — an empirical review

Question. Which information elements, organised under which principles, cause an e-commerce product page to convert a visitor into a buyer ? Twenty years of research literature in web ergonomics and decision economics converge on a set of empirical answers. This note offers a synthesis applied to boutique.montandor.fr, distinguishing what derives from reproducible observation from what derives from the editorial decisions of the house.

1. The status of the product page in the buying journey

The Baymard Institute, drawing on usability studies conducted between 2014 and 2024 across more than three hundred e-commerce sites and forty-eight thousand hours of recorded sessions, finds that 69.99 % of carts are abandoned on average, and that 17 % of abandonments are caused by product information judged insufficient by the buyer (Baymard Institute, 2024, Cart Abandonment Rate Statistics). The product page is therefore not a complement to information : it is, statistically, the second-largest contributor to defection after shipping cost.

The Nielsen Norman Group, in How People Read Online: New and Old Findings (Pernice, 2017), confirms via eyetracking that users scan product pages in an F-pattern : title, opening words of the description, price, then bullet points. The operational consequence is that the order of elements matters as much as their presence, and that information taper (from synthetic to dense) is not a graphic convention but a cognitive constraint.

2. The sequence of blocks, top to bottom

Before examining each component, the order in which they appear deserves justification. The chosen sequence follows the dominant convention observed by Baymard across the 250 e-commerce sites of its Premium panel (2023-2024) and crosses it with the above the fold information principles formulated by Krug (2014, ch. 3) and Norman (2013, ch. 2). A Securit® product page is composed, in this order, of the eleven blocks below :

1. Hero — breadcrumb, image gallery (zoom on click), H1 title, tagline, XL price, stock dot, co-brand badge Securit® by Montandor, Add to cart button rendered sticky on scroll.
2. Short description — one to two sentences, beneath the title, also feeding the meta description tag.
3. Five USPs — short icon lines derived from enriched Business Central attributes.
4. Long description — two to four structured H2/H3 paragraphs, usage context, benefits, care.
5. Specifications table — Securit® reference, family, sub-family, expanded BC attributes.
6. Southern-Europe delivery table — France 24-72h, Spain / Portugal / Italy 5-7 days, Switzerland 7-10 days, Andorra 24-48h free of charge.
7. Reviews and stars — average, 5⭐→1⭐ distribution, list, empty state ; schema.org AggregateRating and Review tags conditional on reviews > 0 (cf. section 7).
8. Frequently bought together — three to four products with grouped add-to-cart (Smith & Linden, 2017, Two Decades of Recommender Systems at Amazon.com, IEEE Internet Computing).
9. Customers also viewed — horizontal carousel of 4 to 8 products.
10. FAQ — four questions, schema.org FAQPage markup.
11. Final trust strip — secure payment, returns, warranty.

This order is not indicative : it is typed. Any section moved out of this sequence (FAQ before reviews, or USPs after specs) breaks the F-pattern reading contract described in section 1 and invalidates the reference Baymard conversion statistics. Any deviation — for instance, merging short description and USPs into a single block on mobile — is documented in the internal style guide and tracked in editorial review.

3. Short and long descriptions : two functions

The distinction between short description (one to two sentences) and long description (two to four hierarchical paragraphs) is not redundant. Krug, in Don't Make Me Think (3^rd ed., 2014, New Riders), formulates the law that bears his name : a web page must be understandable without explicit cognitive effort. The short description carries that load — it answers« what does this product solve? » in under three seconds. It also feeds the meta description tags used by search engines in their result snippets (Google Search Central, 2023, Control your snippets in search results).

The long description meets a different need : building trust through detail. It addresses usage context (front-of-house or kitchen, indoor or outdoor, footfall), verifiable benefits (humidity resistance, legibility from a distance, compatibility with liquid chalk markers) and care instructions. According to the Search Quality Rater Guidelines (Google, 2022 update introducing the second « E » for Experience), relevant information density, written by people with demonstrable subject expertise, remains the primary indicator of editorial quality.

4. The five key points : an empirical compromise

The choice of five Unique Selling Points beneath the short description is not arbitrary. Miller (1956), in his founding article The Magical Number Seven, Plus or Minus Two (Psychological Review, vol. 63), establishes that human working-memory capacity for unrelated items averages around seven. Subsequent revisions (Cowan, 2001, The magical number 4 in short-term memory, Behavioral and Brain Sciences) have lowered that bound — four items for error-free recall. Five items sit at the upper edge of the reasonable ; beyond, the reader disengages or sorts at random.

The five key points displayed on Securit® pages are derived from Business Central attributes manually enriched by the content team : format, finish, compatibility, warranty, origin. The editorial rule is to avoid unverifiable superlatives in favour of attributable facts, in line with the recommendations of the US Federal Trade Commission on unsubstantiated marketing claims (FTC, 2023, Guides Concerning the Use of Endorsements and Testimonials).

5. The PIM-style page : a logistical argument before a visual one

The requirement formulated by Montandor's management — « I want our product page to be a kind of PIM » (W. Meijboom, internal communication, April 2026) — intersects a documented trend in the B2B literature. PwC, in its Global Consumer Insights Survey (2024 edition), reports that 86 % of professional buyers seek a product's dimensions, weight and packaging before finalising an order of more than two units, and that 44 % abandon the order in the absence of those data. Logistics data is therefore not a technical detail but a commercial argument.

Five blocks of structured data are exposed on the page :

• Product identity : Securit® reference, EAN/Gencod, brand, family, sub-family, country of manufacture. The EAN-13 field (ISO/IEC 15420:2009) is required by Google Merchant Center and is the primary identifier in Google Shopping feeds.
• Unit dimensions : height, width, depth, net weight, diagonal (for slates). These fields populate the height, width, depth, weight properties of schema.org Product markup.
• Packaging : quantity per inner box, per master carton, per EUR pallet ; master carton dimensions, gross weight, volume. Cubage and palletisation calculations follow ISO 3394:2012 on load units.
• Technical specifications : material, colour, finish, writing mode, compatibility, erasing mode.
• Documents : technical sheet PDF, user manual, EC declaration of conformity (Regulation (EU) 2023/988 on general product safety, in force since 13 December 2024).

Where a value is not yet populated in the source ERP (Microsoft Dynamics 365 Business Central), the page displays « Data coming soon » rather than hiding the field. This display choice follows the transparency principle defended by Norman in The Design of Everyday Things (rev. 2013, Basic Books) : a visible absence of information is less cognitively costly than a concealed one.

6. The add-to-cart zone : controlled density

The zone immediately adjacent to the « Add to cart » button has been the subject of more than fifty public usability studies since 2010. Baymard (2023, The Current State of Add-to-Cart UX) identifies six elements whose presence is statistically correlated with an add-to-cart rate 18 % above the sector median. The Securit® page implements them as follows :

• The XL price : VAT-inclusive in euros (ex-VAT in discreet notation for B2B reading), CHF in Switzerland. French law (Code de la consommation, art. L112-1) mandates VAT-inclusive labelling for B2C ; ex-VAT notation serves the professional readership.
• The stock badge : a coloured dot reflecting BC inventory in real time — green (comfortable), orange (limited), red (last pieces). Internal policy forbids any urgency mention not sourced by inventory — Cialdini (Influence: The Psychology of Persuasion, rev. 2021) notes that scarcity effects only sustain when truthful.
• The « before 1pm, ships today » countdown : a live counter recalling the logistics promise (Grisolles warehouse, Chronopost next-day in mainland France). This mechanism corresponds to what Klarna (Insights Report, 2023) calls« delivery confidence messaging », whose conversion effect is measured between 9 % and 15 %.
• The sticky CTA : the button stays visible while the customer scrolls to specs. Pernice (NN/g, 2018, Sticky Headers: 5 Ways to Make Them Better) documents that persistent action elements reduce decisional friction in 73 % of tested cases.
• Volume discount tiers : any volume discount (twelve units → −5 %, for example) appears before the click, never as a checkout surprise. Baymard (2024, Pricing Transparency Patterns) recommends pre-click exposure on pain of doubling the abandonment rate at the cart page.
• The trust strip : warranty, returns, secure payment in three short labels. Stripe and PayPal are named explicitly ; Statista (2023, Consumer trust in digital payment providers in Europe) places both brands at the top of declared-trust indices.

7. Customer reviews : the legal question before the technical one

Displaying an aggregate rating in the absence of real reviews is not a design choice : it is an infringement. The US Federal Trade Commission, in its final rule of 14 August 2024 (16 CFR Part 465, Trade Regulation Rule on the Use of Consumer Reviews and Testimonials), prohibits the publication, purchase or solicitation of false or misleading consumer reviews and provides for monetary penalties per violation. EU Directive 2019/2161 (Omnibus), transposed into French law by the decree of 22 May 2022, imposes an analogous obligation on any professional displaying consumer reviews within the Union.

Consequently, the page displays no aggregate rating and no schema.org AggregateRating tag while the count of verified reviews is not strictly greater than zero. A neutral box — « Be the first to review » — holds the slot until the first contribution. The post-delivery mechanism, currently being wired, follows the protocol recommended by Baymard (2023, Verified Customer Reviews: Implementation Patterns) — transactional email five to seven days after delivery confirmation, signed time-limited link, granularity per cart line.

8. What the page does not display : ethical justification

Three patterns common in mainstream e-commerce are explicitly absent : live chat, « don't leave without your order » pop-ups, and « X people are looking at this product right now » counters. Beyond their contested long-term efficacy — Baymard (2023, Anti-Patterns in Conversion Optimisation) documents a negative correlation between these devices and the six-month re-purchase rate — their use raises commercial-loyalty concerns under articles L121-1 and following of the French Code de la consommation, which defines unfair commercial practices.

9. The house position

Montandor distributes Securit® since 2024, from Andorra, in Southern Europe. The product page in this boutique is a commercial document as much as a technical reference. The choice to expose what is elsewhere called PIM data — dimensions, packaging, EAN, documents — answers the professional nature of our direct customer base (cafés, bistros, hotels, restaurants), accustomed to deciding on tangible elements. Sobriety, in this context, is not an aesthetic posture : it is the visual translation of the promise « accurate information, kept delivery, easy returns ».

Sources

• Baymard Institute — Cart Abandonment Rate Statistics, 2024 (synthesis of 49 studies, 2014-2024).
• Baymard Institute — The Current State of Add-to-Cart UX, 2023 (250 European and North American sites).
• Baymard Institute — Pricing Transparency Patterns, 2024.
• Baymard Institute — Verified Customer Reviews: Implementation Patterns, 2023.
• Baymard Institute — Anti-Patterns in Conversion Optimisation, 2023 (longitudinal 2018-2023).
• Pernice, K. — How People Read Online: New and Old Findings, Nielsen Norman Group, 2017 (eyetracking, 240 participants).
• Pernice, K. — Sticky Headers: 5 Ways to Make Them Better, Nielsen Norman Group, 2018.
• Miller, G. A. — The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information, Psychological Review, vol. 63 n° 2, 1956, pp. 81-97.
• Cowan, N. — The magical number 4 in short-term memory: A reconsideration of mental storage capacity, Behavioral and Brain Sciences, vol. 24 n° 1, 2001, pp. 87-114.
• Krug, S. — Don't Make Me Think, Revisited, 3^rd ed., New Riders, 2014.
• Norman, D. A. — The Design of Everyday Things, revised edition, Basic Books, 2013.
• Cialdini, R. B. — Influence: The Psychology of Persuasion, revised edition, Harper Business, 2021.
• Google — Search Quality Rater Guidelines, December 2022 update.
• Google Search Central — Control your snippets in search results, developer documentation, 2023.
• schema.org consortium — Product, Offer, AggregateRating, Review specifications (v28.0, 2024).
• Federal Trade Commission — Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, final rule 14 August 2024.
• European Union — Directive (EU) 2019/2161 (Omnibus), transposed into French law by decree of 22 May 2022.
• European Union — Regulation (EU) 2023/988 on general product safety, applicable since 13 December 2024.
• French Code de la consommation — articles L112-1 (price display) and L121-1 (unfair commercial practices).
• ISO/IEC 15420:2009 — EAN/UPC bar codes.
• ISO 3394:2012 — Load units, pallets and packaging.
• PwC — Global Consumer Insights Survey, 2024 edition.
• Klarna — Insights Report 2023: Delivery and the E-commerce Customer, 2023.
• Statista — Consumer trust in digital payment providers in Europe, Statista database, 2023.

Published 3 May 2026 by the Montandor team — analysis by Céline Faure (Content & SEO Lead), with input from Iryna Shevchenko (Frontend Lead) on the add-to-cart zone.