Skip to main content
Web Development & UX

15 Ecommerce Website Checks Before You Add More Features

Review 15 ecommerce essentials covering products, navigation, checkout, delivery, accessibility and measurement before investing in more store features.

MattDarm8 min read
Illustrated shopping trolley and ecommerce icons on a laptop screen
Illustration accompanying MattDarm's guide: 15 Ecommerce Website Checks Before You Add More Features.

Key Takeaways

  • Make products, total costs and fulfilment understandable before adding optional features.
  • Test the complete purchase and recovery journey, not just the storefront design.
  • Use genuine reviews without incentives or selective positive invitations.
  • Judge success through completed orders, customer experience and commercial value.

An ecommerce website does not need every fashionable feature to sell effectively. It needs to help a customer find a suitable product, understand the terms, pay reliably and receive what was promised. Extra recommendations, chatbots or loyalty mechanics cannot compensate for missing delivery information or a broken checkout.

Use these 15 checks before planning another plugin or redesign. They are acceptance questions, not guaranteed conversion improvements. Prioritise according to your products, customers and operating constraints, and test changes in an appropriate environment before they affect live orders.

1. A usable mobile shopping journey

Open a category, compare products, choose a variation and proceed through checkout on a phone. Check menus, filter controls, keyboards and validation messages. Do not stop at whether the homepage fits the screen.

Use realistic product names and images. Long titles, out-of-stock variations and promotional messages can expose layout problems that an empty template does not show. Ensure sticky elements do not cover prices, delivery information or the payment action.

2. Product information that supports a decision

Explain dimensions, materials, compatibility, included items and any limitations relevant to the purchase. The right information depends on the product; generic manufacturer copy may leave important questions unanswered.

Use accurate images and show meaningful detail. If an image is a rendering or a proposed configuration, label it appropriately. Keep informative alt text and do not use decorative images as a substitute for essential written specifications.

3. Categories and navigation customers understand

Organise products around how customers search and compare, not only how the warehouse or supplier catalogues them. Use clear category labels and provide a route back to the wider selection.

Test a few realistic tasks with people unfamiliar with the store. The navigation guide explains how to observe confusion without coaching. There is no universal ideal number of menu items for every catalogue.

4. Search and filters suited to the catalogue

If the catalogue needs search, test common product terms, model numbers, spelling variations and no-result cases. Useful filters should reflect meaningful buying criteria and the actual product data available.

Do not add AI recommendations before basic search works. Check empty combinations, unavailable stock and how filters behave on mobile. A clear no-results message with a recovery route is better than a silent blank page.

5. Clear price and total-cost information

Make the price and applicable charges understandable before the customer commits. Explain delivery costs and any conditions that affect the total. Avoid surprising mandatory additions late in checkout.

Check the requirements relevant to the business and products. The UK government's distance-selling guidance is an official starting point, not a replacement for advice about a particular legal situation. Keep the website aligned with the actual terms you can fulfil.

6. Delivery, returns and support information

Explain dispatch expectations, coverage, collection options and what happens if a product is unsuitable or a delivery fails. Do not present an estimate as a guaranteed date unless the operation supports that commitment.

Make support easy to find before and after purchase. A chatbot can help locate approved information, but it should not invent policy exceptions or prevent access to a person when the issue needs judgement.

7. A checkout that collects the right information

Review every field and its purpose. Offer a suitable guest route where the business model permits it, and avoid demanding an account without a clear reason. Keep required and optional information understandable.

Use visible labels, appropriate input types and useful errors. The W3C forms tutorial provides practical guidance. Preserve entered details after an error so customers do not have to repeat the whole task.

8. Reliable payment and order confirmation

Use reputable payment infrastructure and test the supported methods in the appropriate test mode. Check successful, declined, cancelled and interrupted payments. A success screen should correspond to a real accepted transaction and order state.

Verify confirmation emails, stock changes and fulfilment notifications. Repeated callbacks or retries must not create duplicate orders or charges. Keep payment-provider access controlled and do not collect card details through an improvised form.

9. Performance on the pages that matter

Measure category, product and checkout behaviour with realistic content. Images, scripts and third-party widgets can affect different pages differently. A strong homepage score does not establish that the purchase journey is responsive.

Use Web Vitals to understand loading, responsiveness and stability. Do not translate each second into an assumed percentage of lost revenue. Improve measured bottlenecks and review the business effect in your own context.

10. Accessible interactions and clear states

Check keyboard access, visible focus, product-option labels, error messages and status updates. Do not rely on colour alone to distinguish availability or an invalid selection. Test the important task with relevant assistive technology where possible.

Accessibility is not just a final visual adjustment. A product selector, filter drawer or payment step can introduce barriers even when the text colours look acceptable. Include these interactions in the build's acceptance criteria.

11. Genuine reviews and accurately labelled proof

Display authentic customer feedback and preserve its meaning. Do not manufacture reviews, offer rewards for Google reviews or ask only customers expected to be positive. Follow the applicable platform rules and relevant legal requirements for any review programme.

Use the Google review-request guide for a neutral process. If you show customer images, obtain appropriate permission. Do not present a stock image or a drafted quote as a real customer's experience.

12. Crawlable products and useful search information

Check that intended indexable products and categories have stable URLs, accurate titles and useful descriptions. Keep important links crawlable and ensure discontinued or moved products are handled according to their actual replacement and business context.

Google's ecommerce guidance covers site structure and product discovery. Structured data should match the visible product and current offer. It is not a guarantee of a rich result or an excuse to publish unavailable prices.

13. Recovery messages with appropriate permissions

A basket reminder can be useful in some circumstances, but it is not automatically permitted simply because someone entered an email address. Review the purpose, consent or other applicable conditions, and the rules for the channel before automating messages.

Use the ICO's direct-marketing guidance as a starting point. Keep frequency restrained, provide appropriate choices and avoid making an unverified claim that a particular product is about to sell out.

14. Optional features justified by a real need

Personalisation, loyalty schemes, subscriptions, chat and mobile apps can be worthwhile, but each introduces cost and maintenance. Define the problem before adding the feature. A confusing catalogue does not necessarily need an AI recommendation engine; it may need better product data.

Read the mobile-app decision guide before adding another customer platform. Test a bounded version of the proposed feature and measure meaningful use. Do not treat the presence of a feature as proof that it improves retention.

15. Measurement connected to orders and margin

Track successful purchases and important diagnostic steps without duplicating events. Reconcile analytics with actual order records and recognise consent or attribution limitations. A payment-page visit is not the same as a completed sale.

Review returns, support costs, fulfilment and margin as well as gross revenue. A promotion that increases orders but creates unprofitable fulfilment may not be a commercial improvement. Keep test orders distinguishable and remove or cancel them through the correct process.

Turn the checklist into a release plan

Record each problem with its affected page, evidence, owner and acceptance test. Fix broken purchases, misleading information and inaccessible interactions before optional merchandising features. Test on staging and plan carefully around live stock and order data.

After release, recheck representative products and the full customer journey. Keep a dated change record so later reports can distinguish a content update, price change, promotion and measurement fix. Do not assume every variation in revenue came from the design work.

If visits are not leading to useful actions, use the traffic-without-enquiries guide as a diagnostic model. MattDarm can help with ecommerce development, SEO services and conversion rate optimisation. Share the store and the customer problem so the work starts with a verified priority.

Frequently Asked Questions

Does every store need all 15 features immediately?

No. These are checks to prioritise according to the business. Reliable product information, payment and fulfilment are more fundamental than optional personalisation, loyalty or AI features.

Should we add a chatbot to improve sales?

Only when it addresses a real customer need and can use accurate information with a human handoff. Fix missing product details and support routes first rather than expecting a chatbot to compensate for them.

Can we reward customers for Google reviews?

No. Google's policy prohibits incentives for reviews. Use a neutral invitation and do not selectively request positive feedback or write reviews for customers.

What should a checkout test include?

Test successful and failed payments, validation, interruptions, confirmation messages, emails, stock changes and duplicate prevention. Use the appropriate test environment and verify the receiving systems, not only the visible page.

How do we measure a store improvement?

Compare reliable order and customer-journey evidence over suitable periods, including returns, support and margin. Record other changes and measurement limitations instead of attributing every revenue movement to one feature.

Share this article

Stay ahead of the curve

Weekly insights on web development, AI, branding & digital marketing. No spam, unsubscribe anytime.

By subscribing you agree to our Privacy Policy. Unsubscribe at any time.

Adam Saez
Alina Stefanovičiūtė
Daniel Ashby
Matt Laybourn
Richard Jones
Paul Campbell

People we've worked with

Real projects, built together

Let’s Grow Your Business Together

Tell us about your project and we’ll show you exactly how we’d grow your business. Book a free 30-minute discovery call, no pressure.