Why Fashion Product Views Do Not Become Add-to-Carts

Fashion shoppers view products without adding to cart when the page leaves a question open: fit, fabric, true color, price against what they already own, or whether the photo resembles what arrives. Before redesigning anything, split view-to-add-to-cart by device and traffic source in Shopify. The gap is rarely spread evenly, and the segment where it collapses tells you which question to answer first.

Views on the left, add-to-carts on the right, and a long night of guessing in between. Editorial image in Classic Antla disposable-camera style.

When a linen shirt collects thousands of views and a trickle of carts, the first instinct is to reshoot it. Reshoots are slow, expensive, and only useful if the photos were the problem. Sometimes they were. Just as often the photos are fine on a laptop and useless on a phone, or the shoppers arriving from one ad were never looking for a linen shirt.

The job below is diagnosis first, then one change at a time.

Diagnose by device and source

A store-wide add-to-cart rate is an average of very different visits. A returning email subscriber on a laptop and a first-time Instagram visitor on a phone are not doing the same thing, and averaging them hides both.

Shopify gives you the session-level view out of the box. The Behavior reports include a Conversion rate breakdown funnel that runs from every session through cart additions and checkout to completed checkout, plus Sessions by device and Sessions by landing page. Run the breakdown for mobile and desktop separately, then for your main traffic sources.

For product-level detail you need events. Shopify’s standard pixel events include product_viewed and product_added_to_cart, so any analytics tool fed by them can compute view-to-cart per product, per device, and per source. That is the number this article cares about.

Before you read it, filter out bots. Price-comparison scrapers and social link previews request product pages all day and never add anything to a cart. Shopify’s bot filtering labels each session as human or bot, and reports that support the filter let you exclude the bot sessions. Skip this step and a scraper can make a perfectly good dress look like a conversion problem.

Then look for the collapse. Here is what I would suspect first in each pattern:

Where view-to-cart collapses	Suspect first	Quick check
Mobile only	Size selector, gallery, or fit notes that work on desktop but hide on a phone	Open the product on a mid-range phone and try to pick a size one-handed
One paid campaign only	Ad promise and page do not match (color, price, offer, or the look)	Put the ad and the landing page side by side
First-time visitors, all devices	Trust and risk: returns, shipping, sizing system	Can a stranger find the returns policy without leaving the page?
One product type (say, trousers)	A fit question specific to that type	Read that type’s reviews and support tickets for repeated questions
Everywhere, all at once	Price, stock depth, or measurement changes	Check sold-out sizes and recent tracking changes before blaming the page

If nothing collapses and the rate is simply low everywhere, the product itself may be the issue: price, fabric, or a style your audience browses but does not buy. That is a merchandising conversation, not a page redesign.

Audit imagery, sizing and price

Once you know the segment, read the product page as that shopper. Same device, same entry point, same level of knowledge about your brand (usually none).

Baymard’s product page benchmark found that only 49% of ecommerce sites have product page UX rated “decent” or “good.” Fashion pages carry extra weight, which is why Baymard’s apparel research treats conveying fit and feel, including size guides, as its own topic. A generic page audit misses most of it.

For fashion, I audit three things in this order.

Imagery. Does the first image establish shape, or is it a moody crop of a sleeve? Is there a back view, a fabric close-up, and at least one image that shows how the garment moves or drapes? Does the color swatch change the main image, or does the shopper have to trust that “sage” means what they think it means? If you use variant images or combined listings, check they behave on mobile. The variant images and combined listings breakdown covers the common failure modes.

Sizing. Is the model’s height and size stated next to the photos? Are garment measurements available alongside a body chart? Is there a fit note (“cut close through the shoulder, size up if between sizes”) near the size selector, where the decision happens? Sold-out sizes should be visible and marked, not silently removed, so a shopper does not assume the item runs in odd sizes.

Price. Price objections on a product page rarely look like price objections. They look like a shopper zooming into the fabric, checking the composition, and leaving. The question is whether the page justifies the number: fabric weight, construction details, and care.

Shipping thresholds and return costs belong here too, because a shopper doing mental math on a $95 shirt is also doing it on the $8 return label. In the NRF’s 2024 returns research, 76% of consumers said free returns are a key factor in deciding where to shop.

Photo reviews help with all three when they show real bodies and real color. The photo reviews and social proof guide explains how to make reviews answer fit questions rather than just add stars.

Test preview and confidence cues

Confidence cues are small pieces of evidence that answer the specific question the shopper is holding. They are not badges. “Free returns” in a footer is policy. “Free returns within 30 days, and exchanges ship before you send yours back” beside the size selector is a confidence cue.

Useful cues for fashion, roughly in order of cost:

1. A one-line fit note beside the size selector
2. Model height and size worn, beside the gallery
3. Garment measurements for the selected size
4. Reviews filterable by fit (“runs small,” “true to size”) and by height
5. A short video of the garment moving
6. A personal preview that shows the piece on the shopper

Test them in the segment where view-to-cart collapsed, not store-wide. A fit note that rescues mobile Instagram traffic may do nothing for desktop email subscribers who already know your sizing.

If the open question is appearance (how this shape, color, or length looks on me), a virtual try-on is one cue worth testing. Across Antla merchants, shoppers who use try-on convert about 35% higher on average than shoppers on the same stores who do not. That compares people who chose to use the feature with people who did not, so it describes an association rather than proving the preview caused the difference. Antla merchant reporting, cohort dates not published.

Try-on shows how a piece may look on the shopper. It does not measure their body or confirm fit, so keep your measurements and fit notes. If you want to run that test on the segment where views stall, Antla puts try-on on Shopify product pages.

For the broader set of fit signals and how to measure them, see fit confidence metrics and PDP tactics.

Measure add-to-cart lift

Measure the change where you made it. If you added a fit note for mobile shoppers on trousers, the metric is view-to-cart for mobile sessions on trouser pages. Store-wide conversion will move too slowly and too noisily to tell you anything useful for weeks.

Use one worksheet per test. The first row below is illustrative, not a benchmark:

Field	Example (illustrative)	Your test
Segment	Mobile, paid social, wide-leg trousers
Question you think is open	Does the rise sit at the waist or the hip?
Change	Rise measurement and model height beside size selector
Baseline view-to-cart	3.1% over the prior 14 days
Test window	14 days, same campaigns running
Result view-to-cart	Fill in after the window
Downstream check	Reached checkout and completed checkout for the same segment
Later check	Size-related returns for orders in the window

Three rules keep the readout honest.

First, compare against a control. A holdout or an A/B split is best. A before-and-after comparison is acceptable only if traffic mix and promotions stayed flat, which they rarely do.

Second, follow the cart downstream. Some changes raise add-to-cart without raising orders. A countdown timer is the classic example: more carts, same purchases, slightly more tired shoppers.

Third, wait for enough carts. A segment with a few dozen add-to-carts a week can swing on one good afternoon. Extend the window rather than calling a winner early.

For a finer-grained view of what shoppers do between the view and the cart, see product page engagement as conversion quality.

Where this diagnosis stops working

View-to-cart tells you where shoppers stop, not why. The why comes from reading the page as the shopper, from reviews, and from support questions. If you skip those, you are testing guesses with better formatting.

A few limits to keep in mind:

• Consent gaps. Shoppers who decline analytics tracking may not appear in third-party tools, and the missing share can differ by region and device.
• Cross-device journeys. A shopper who views on a phone and buys on a laptop looks like a mobile failure and a desktop success.
• Small catalogs and slow sellers. Products with low traffic need longer windows or grouping by product type.
• Add-to-cart is not purchase. It is a leading indicator. Keep checking orders and returns for the same segment.

If sizing looks like the culprit, why size charts fail on Shopify fashion shows the measurement and fit gaps that show up in events and return reasons. For the full PDP checklist once diagnosis is done, see Shopify PDP conversion optimization for fashion.

Start with one segment this week

Open the Conversion rate breakdown in Shopify, split it by device, and filter to human sessions. Pick the worst segment with real volume. Visit five product pages the way that shopper arrives, on the same kind of device, and write down every question the page leaves open.

Choose the cheapest cue that answers the most common question, ship it for that segment only, and fill in the worksheet. One clean test is worth more than a redesign built on an average.

---

About the author: Aaron is the founder of Antla and spends an unreasonable share of his week reading fashion product pages on a phone.