Why is your online store not selling? What to check first

If your online store is not selling, do not begin by increasing the advertising budget or replacing the platform. First identify where shoppers leave: before reaching a product, on the product page, after adding an item to the basket or during payment. Each failure point calls for a different fix.

A useful diagnosis combines analytics, test orders and a review of the offer. A conversion rate alone cannot explain the cause. The sequence below separates a traffic problem from a store problem.

1. Confirm that measurement reflects real orders

Place a test order before drawing conclusions from reports. Complete it on a phone and a computer. It should appear in the store administration, payment provider and analytics without being counted again when the confirmation page is refreshed.

A practical ecommerce setup records at least these stages:

• viewing a product,
• adding a product to the basket,
• beginning checkout,
• completing a purchase with its transaction ID and value.

Google's ecommerce measurement documentation describes the view_item, add_to_cart, begin_checkout and purchase events. Recording only purchases hides where potential customers leave. Missing purchase events can also make a successful campaign look ineffective.

Reconcile GA4 with the store and payment records. A discrepancy is a measurement issue, not evidence that the offer is wrong.

2. Separate insufficient traffic from poor conversion

A store with no orders can have two very different problems:

• too few relevant people reach its products,
• people view the offer but do not finish buying.

Review traffic sources, devices, landing pages and viewed products. A broad informational query does not carry the same intent as a search for a specific model. An advert promising one product should not send the prospect to a homepage where the search starts again.

For paid campaigns, compare the advert with the destination page's heading, price and availability. The guide to choosing a landing page or website page for Google Ads explains this relationship. In ecommerce, the destination will often be a specific product or focused category.

3. Locate the weakest step in the buying journey

You do not need an elaborate dashboard. Compare consecutive steps for the same period, traffic source and device type:

• few product views point towards acquisition, navigation, category or search problems,
• many views but few basket additions point towards the product, price, content or availability,
• basket additions without checkout starts bring attention to the basket, delivery costs or purchase terms,
• checkout starts without purchases require form, payment and technical testing,
• orders in the store but no purchases in analytics indicate broken measurement.

Compare like with like. Combining every channel, customer type and device may conceal the cause.

4. Make the product page answer buying questions

A product page should help a shopper decide whether the item fits and what happens after the order. Check whether it shows, without a search through terms and conditions:

• the price and an honest presentation of any reduction,
• available variants, sizes or configurations,
• current stock status,
• the delivery time and cost, or a clear way to calculate them,
• return information relevant to the decision,
• images that show the item, scale and important details,
• specifications needed for comparison,
• an obvious purchase button and visible feedback after it is used.

The description should remove uncertainty, not repeat the product name. A technical item needs dimensions and compatibility, clothing needs material and sizing information, and a made-to-order item needs a clear process and lead time.

The page, product advertising and feed must agree. The Google Merchant Center product data specification requires titles and descriptions to match the landing page. Conflicting prices or availability undermine both the campaign and shopper confidence.

5. Reveal delivery costs and availability early

Do not require all customer details before revealing a material surcharge. If delivery depends on basket value, destination, parcel size or method, explain the rule and allow calculation before the final step.

Test edge cases as well as the standard order:

• a product appears available but no variant can be selected,
• free delivery is advertised but not applied in the basket,
• a discount code changes the price differently from the advert,
• the stock system accepts an order for an unavailable item,
• delivery estimates differ between the product page and order summary.

Google can use product structured data to show price, availability, delivery and return information. Its merchant listing documentation explains the requirements for Product and Offer. Markup cannot repair inconsistent customer-facing information.

6. Complete the purchase on a real phone

A desktop responsive preview is only the beginning. On a real phone, check:

• filtering and sorting a long product list,
• selecting variants and changing quantities,
• how the keyboard affects checkout fields,
• error messages and where focus moves after an error,
• hand-off to a bank, digital wallet or card authentication,
• return to the store after payment,
• whether a banner or floating element covers the purchase button.

Assess performance with real-user data, not only one laboratory run. Current Core Web Vitals guidance recommends LCP within 2.5 seconds, INP at 200 milliseconds or less and CLS at 0.1 or less at the 75th percentile, split by mobile and desktop. These metrics reveal slow loading, delayed interactions and unstable layouts, but cannot guarantee sales.

7. Remove friction from the basket and checkout

The basket should state products, variants, quantities, reductions, delivery and the final amount. Ask only for information needed to fulfil the order or meet the business's obligations. If the model does not require an account, forcing its creation adds an unnecessary step.

The W3C accessible forms tutorial recommends visible labels, instructions, helpful validation and clear error and completion notices. These principles matter because an unclear checkout error can stop a ready buyer.

Test an invalid postcode, a missing agreement, a declined payment and a second attempt. “Something went wrong” does not say whether to correct a field, choose another method or return later. Confirmation should provide an order reference and next steps. The same principles appear in the guide to website contact forms and better enquiries.

8. Check product visibility before buying more traffic

If the journey works but too few people reach it, review indexation, the product feed, category structure and incoming queries. Structured data helps Google understand a product, but does not replace an accessible page, accurate title, useful description, suitable image and internal links.

For a larger catalogue, inspect filters that generate similar URLs, discontinued items and variants. The guide to technical SEO for a new website covers indexation controls, canonicals and sitemaps. Fix discovery and purchase blockers before scaling paid acquisition.

A repair order that reduces guesswork

A practical plan looks like this:

1. Place and document test orders on mobile and desktop.
2. Verify ecommerce events and reconcile revenue with store records.
3. Identify the weakest journey stage for the main traffic sources.
4. Fix payment, delivery, variant and checkout errors.
5. Improve the most viewed product pages and important categories.
6. Compare the same data after release before starting the next change.

Avoid changing prices, advertising, product layouts and checkout together if you want to learn what worked. Fix urgent defects immediately, but run business experiments so they can be evaluated.

When are focused fixes enough, and when should you rebuild?

Targeted work is often enough for measurement, product content, delivery configuration, one checkout step or an integration defect. A rebuild is more reasonable when the platform blocks essential payment and delivery methods, catalogue maintenance is unreliable and mobile use depends on workarounds.

The comparison of WooCommerce and Shopify covers maintenance, control and integrations. If the broader question is whether to repair the store or start again, use the website redesign decision guide.

AMCompany designs and develops online stores, including analytics, buying journeys, performance and technical SEO. Review the services and portfolio. To begin with a diagnosis, describe the store and the problem, including the platform, traffic sources and stage where purchases most often stop.