Contents

A product page that receives traffic but does not generate orders may have problems with its presentation, offering, usability, or functionality. However, the lack of sales does not automatically prove that the page is the cause. First, you need to check whether the users you attract are the right ones and at what stage they abandon.

If visitors land on a product but leave without adding it to their cart, the investigation should focus on the relevance of the traffic, product, offer, and page. If the product is added to cart, but order not completed, the problem may occur later, in the cart, checkout, payment, or delivery.

Where does the user get lost?

Before we modify the page, we need to determine what happens after access.

Observed behaviorCauses worth checking out
Users leave the page quicklyInappropriate traffic / slow loading / message different from the advertisement / product immediately perceived as irrelevant
I am viewing the product but not selecting variantsInsufficient information / difficult to understand options / lack of desired size or color
Select options, but do not add to cartPrice / unknown costs / lack of reliability / hard-to-see button
Add the product to cart, but do not purchaseShipping / payment / account required / errors or difficulties in checkout
Orders appear but are canceledIncorrect stock / delivery time / delayed confirmation / product presented improperly

Page conversion rate should be analyzed along with cart additions, checkout starts, and completed orders. On its own, it doesn't tell you where the bottleneck is.

We always check the product page first: it decides whether the user trusts it. We look at how the information is presented, what elements support trust in the brand, and whether the important data is arranged in a way that is easy to understand. Only then do we discuss the price and the checkout process.

Is the traffic reaching the right product?

A page can be well-built and still not convert if visitors came with a different intention.

Let’s say a user searches for a product for a specific app, but the ad sends them to a similar variation for a different purpose. The person can view the images and description, then leave. The page didn’t fail to showcase the product. The campaign created an inappropriate association between the search and the offer.

They need to be compared:

  • the terms used by users;
  • the advertising message;
  • the product they are sent to;
  • the price or benefit mentioned before the click;
  • availability of the searched variant;
  • the stage the user is in.

„"Relevant traffic doesn't just mean people interested in the general category. It means people for whom the product, price, terms, and timing of purchase are compatible."”

If the differences between campaigns are large, the page should be analyzed separately for each source. An overall rate can hide the fact that organic traffic is buying, while a campaign is sending users with weaker intent.

What should the user understand from the first screen?

The first screen should quickly answer the questions that decide whether the person continues:

  • What product is it?
  • For whom or for what use is it suitable?
  • How much does it cost?
  • Is it available?
  • Which option can I select?
  • What do I need to do to buy it?

Not all details need to be crammed at the top. However, information that immediately influences the decision shouldn't be hidden behind multiple sections, tabs, or paragraphs.

Depending on the product, the first screen may contain the name, images, price, variants, availability, a summary of benefits, and a buy button. For a configurable product, the user also needs to understand which choices are mandatory.

Primul ecran al unei pagini de produs pe telefon: titlul produsului, eticheta de vârstă, starea stocului, fotografia, prețul vechi tăiat, prețul curent și butonul de adăugare în coș.
The first screen of a product page, on a phone. In the order they appear: what the product is, what age it is for, if it's in stock, what it looks like, how much it costs now vs. how much it cost, and the add to cart button. The questions above are answered without any scrolling. The screenshot is from Toy World store, built by us.

A visually appealing first screen is not enough if the person cannot identify the product or the next step.

Does the page describe the product or just praise it?

General texts like "high quality", "modern design" or "the perfect choice" provide little information for a decision.

The description should help the user determine whether the product solves their need. This may require explaining:

  • materials and dimensions;
  • how to use;
  • compatibility;
  • package contents;
  • the differences between the variants;
  • maintenance;
  • product limitations;
  • situations in which it is recommended;
  • situations where it is not appropriate.

Not all products require the same amount of information. For a simple item, a short description may be enough. For technical, customizable, or high-priced products, a lack of detail may force the user to look elsewhere for answers.

Nielsen Norman Group includes among the necessary elements of an ecommerce page a descriptive name, recognizable images, price, options, availability, the ability to add to cart, and an informative description. Nielsen Norman Group recommendations for product pages1

The role of the description is not to take up space or introduce repeated SEO phrases. It should reduce questions that hinder the choice.

Images replace in-store viewing

Images are not just for aesthetic purposes. They replace some of the inspection that would be possible in a physical store.

The user may need to see:

  • the product from multiple angles;
  • material and finish details;
  • size in relation to familiar objects;
  • the product during use;
  • included accessories;
  • differences between colors and variants;
  • the packaging or the assembly method.

A beautiful photo can be insufficient if it doesn't show the size or function. At the same time, many nearly identical images can make it difficult to examine without providing new information.

For some products, a short video or 360-degree view can reduce uncertainty. For others, a well-framed photo and close-up detail are more helpful. The choice of format should be based on the buyer’s real questions, not the number of materials available.

Baymard’s research treats images, video, variations, the purchase section, shipping, and returns as distinct areas of the product page experience. Its benchmark shows that a significant portion of the stores analyzed still have usability issues on these pages, especially on mobile. Baymard research on product pages2

Are the options easy to select?

Sizes, colors, quantities, and customizable options can create bottlenecks when the relationship between selection, availability, and price is not obvious.

Problems often arise when:

  • it is not visible which variant is selected;
  • unavailable variants can be chosen;
  • changing the option does not update the image;
  • the final price only appears in the cart;
  • the user does not know which fields are required;
  • the variant names are difficult to understand;
  • errors are displayed away from the problematic selection;
  • the choice made is lost after an error.

A custom product needs a logical order of decisions. If the price changes, the update should be visible before adding to cart. If production requires text, a size, or a file, the page should explain the format and when the information will be used.

Not every complex product needs a sophisticated configurator. Sometimes better naming, grouping options, and removing fields that are not needed at that stage are enough.

The user evaluates the final amount, not the displayed price

The user does not only evaluate the price displayed next to the product. He tries to anticipate the final amount and the terms of the purchase.

Uncertainty may arise due to:

  • unknown transport;
  • a minimum quantity;
  • additional fees;
  • options that modify the price;
  • discounts that are difficult to verify;
  • a threshold for free delivery;
  • return costs;
  • differences between the price in the advertisement and the one on the page.

If shipping can only be calculated after entering the address, the page can explain the relevant rule or range. If the price depends on the configuration, the user must see the update before continuing.

Hiding the costs doesn't make them easier to accept. It just moves the moment when the person can give up.

Delivery and returns decide before checkout

Shipping and returns information influences the decision before checkout. If it is only available on a separate and hard-to-find page, the user may postpone the purchase or leave the store to search for the answer.

The product page doesn't have to reproduce the store's policies in full. It can display a relevant summary and a link to the full terms.

Depending on the category, the buyer may need to know:

  • whether the product is in stock;
  • how long it can be shipped;
  • to which areas is it delivered;
  • from what value is shipping free;
  • whether the product can be returned;
  • whether there are exceptions for personalized products;
  • how the warranty applies.

Messages must match the operational process. A promise of fast delivery doesn't help if the inventory or supplier can't support it.

In a store I worked at, information about cost and delivery time existed, but it wasn't highlighted on the product page. The user had to look for it right before making a purchase decision.

We added icons and short texts for the delivery cost, the threshold above which shipping becomes free, and the right to return. We also implemented an estimated delivery system, which shows on the page the interval in which the customer can receive the order.

For some products, sales during the campaign increased by 120%. The information existed before; the difference was in where it appeared and the way it was read.

+120% campaign sales on some products

Secțiunea de livrare dintr-o pagină de produs: costul transportului, pragul de livrare gratuită, dreptul de retur și banda cu intervalul estimat de livrare.
The three conditions that mattered to the buyer, written briefly and accompanied by icons, and below them the estimated delivery interval. The same information was also available before, on the policies page.

Trust comes from verifiable information, not badges

Trust is not built by repeatedly displaying lock icons or texts like „secure payment.” The user looks for clues that the store and the product live up to the promise.

These clues can come from:

  • company details and contact details;
  • accessible policies;
  • verifiable product information;
  • authentic reviews;
  • photos taken by customers;
  • answers to frequently asked questions;
  • warranty information;
  • known payment methods;
  • consistency between ad, page and checkout.

Reviews can be especially helpful when they explain usage, size, compatibility, or the difference between expectations and the product. A collection of reviews without details provides less context.

If there are no reviews yet, they shouldn't be faked. An accurate description, relevant photos, and well-explained conditions are more credible than artificial validation elements.

Does the buy button work as part of the process?

A large button can't make up for a lack of information. However, the main action should be easy to find and use.

Verification includes more than just button color:

  • appears after the user has the necessary information;
  • it is differentiated from secondary actions;
  • indicates what is coming next;
  • works after selecting variants;
  • provides confirmation after adding to cart;
  • keeps the chosen configuration;
  • displays errors next to fields that need to be corrected.

On mobile, the button may disappear under long sections or be covered by pop-ups, banners, and fixed elements. In other situations, it is always visible, but takes up too much of the screen and prevents product review.

The right position must be evaluated in the context of the page. There is no single solution that works for all products.

How do we check the mobile experience?

A page that looks good on desktop can become difficult to use on a small screen.

Images can take up multiple screens, variants can require horizontal scrolling, and product information can be hidden in too many sections. Newsletter, cookie consent, and chat windows can simultaneously cover a significant portion of the content.

Testing should be done on different devices and sizes, not just by shrinking the browser window. Real actions should be performed:

  1. Accessing the page from the advertisement or from a category.
  2. Examining images.
  3. Variant selection.
  4. Reading delivery information.
  5. Adding to cart.
  6. Return to modify selection.
  7. Proceed to checkout.

If an action requires precision, multiple attempts, or repeated returns, friction can affect the decision.

Does the page load and respond well enough?

Technical performance influences the usability of the page, but a good score in a tool does not guarantee conversions.

Google uses Core Web Vitals to measure three components of the experience: loading, interaction responsiveness, and visual stability. The metrics should be verified against user experiences, separately for mobile and desktop. Google documentation on Core Web Vitals3

On a product page, it is worth watching for situations such as:

  • the main image appears with a delay;
  • buttons respond slowly;
  • the page changes position during loading;
  • selecting a variant locks the interface;
  • the gallery consumes too many resources;
  • tracking, chat, and personalization scripts slow down the page;
  • adding to cart does not provide an immediate response.

Technical optimization must be related to the user experience. Reducing a few milliseconds is not necessarily a priority if the product is insufficiently explained or the desired variant cannot be selected.

Is the problem the page or the product?

Not every page that converts poorly needs to be redesigned.

Sometimes users understand the product and decide it's not for them. The cause may be price, lack of demand, competition, product reviews, low differentiation, or commercial terms.

This distinction is important:

SignalPossible interpretation
Users repeatedly search for missing informationThe page may not sufficiently explain the product
Many users check the delivery and then leaveTerm or cost may limit purchase
The popular variant is unavailableStock, not design, may be the main problem
The product is often added to the cartThe page probably fulfills an important part of its role
The page behaves well from one source and poorly from anotherTraffic relevance must be verified
Similar products convert, but this one doesn'tThe offer or product may be more important than the overall structure

Design can make information easier to find. It cannot transform every product into a competitive offering.

When the page isn't the cause, it's usually because the price isn't competitive or the product doesn't offer enough of a reason to buy. In one such case, we increased sales of a product that was above market price by adding a gift with purchase, without changing the page structure.

What data do we use for analysis?

To understand the page, we combine quantitative data with behavioral observations.

Quantitative data can show:

  • product views;
  • additions to cart;
  • variant selections;
  • checkout starts;
  • orders;
  • performance on devices;
  • traffic sources;
  • products and categories;
  • loading times;
  • technical errors.

Session recordings, user testing, and support questions can all explain why certain results appear. They should not be interpreted in isolation. A single unusual recording does not represent the behavior of all visitors.

Internal searches can also provide clues. If users on a product page frequently search for terms like „sizes,” „compatibility,” or „delivery,” the information may be missing or hard to find.

Priority is given to affected users and proximity to decision

Not all observed problems have the same effect.

An intervention should be evaluated after:

  • the number of users affected;
  • bringing the issue closer to the purchasing decision;
  • possible commercial impact;
  • the available evidence;
  • implementation effort;
  • risk of affecting other products or devices.

A payment method that doesn't work takes precedence over changing an icon. A variant that can't be selected takes precedence over rewriting a subtitle. Lack of compatibility information can be more important than a visual change to the button.

For changes with uncertain effects, testing can be useful. However, obvious bugs should not be kept just for the sake of testing. If the button doesn't work on certain phones, the problem needs to be fixed.

Frequently asked questions

What is a good conversion rate for a product page?

There is no universal value. The rate depends on category, price, traffic source, device, product type, and how the conversion is calculated. Comparisons to your own history and similar products are usually more useful than a general average.

Does a low rate mean the page needs to be rebuilt?

No. First, you need to check your traffic, product, inventory, offer, and the stage where users abandon. Sometimes, one-off changes are enough. Rebuilding is justified when the existing structure prevents you from solving important problems.

Should more text be put on the product page?

Only if the text answers relevant questions. Length is not a goal. A simple product may need little explanation. A technical or customizable product may require details, examples, and instructions.

Do reviews automatically increase conversion rates?

Not automatically. Reviews can reduce uncertainty if they are authentic and provide useful information. They do not compensate for a bad product, an uncompetitive price, or a flawed purchasing process.

Is it enough to track the addition to cart?

No. Add to cart shows that the page convinced the user to continue, but does not confirm the sale. Checkout start, payment, confirmed orders, cancellations, and returns should also be tracked.

Can we optimize the page without A/B testing?

Yes. Technical errors, misinformation, and obvious usability issues can be corrected directly. For changes to the message, structure, or presentation with uncertain effect, testing reduces the risk of a conclusion based solely on preferences.

The page must reduce uncertainty before purchase

A product page shouldn't convince every visitor to buy. It should help the right person understand the product, evaluate it, and move on without unnecessary hurdles.

When a page receives traffic but doesn't generate orders, the analysis starts with where users abandon. Then we check traffic relevance, information, images, variations, price, delivery, trust, mobile experience, and technical operation.

The appropriate change depends on the observed cause. Sometimes the page needs to be improved. Other times the campaign, offer, stock must be corrected or the stage following adding to cart.

Sources cited

  1. Nielsen Norman Group recommendations for product pagesnngroup.com
  2. Baymard research on product pagesbaymard.com
  3. Google documentation on Core Web Vitalsweb.dev

The right service for this situation

Conversion optimization and digital performance

Leave a comment

Not published.