Ecommerce best practices that still matter

The ecommerce best practices that still matter are the ones that remove a reason not to buy. Show the full cost early. Answer product questions on the page. Keep checkout short. Prove you're a real business. Work properly on a phone. Do those first. Most of the rest can wait, and some of it can wait for years.

Eugiene Jonathan PisarevskyEcommerce15 min read

A long list makes item 30 look as important as item 1. You have a store to run, so here are 12 practices, in the order I'd do them. Each comes with a test that takes a few minutes. Near the end, there's a list of what you can safely ignore.

Where to start if you only have an hour

Your own numbers will tell you which part of this list to read first. The menu steps in this guide are Shopify's. On another platform, look for the same numbers under different names.

In your admin, go to Analytics, then Reports, and set the Category filter to Behavior. Open Conversion rate breakdown. That report doesn't split phones from computers. So ask Sidekick, Shopify's assistant, to "Show my conversion rate by device type for the last 30 days." Then compare with Littledata's numbers for 438 Shopify stores:

What you're looking atTypical storeIf you're well below it, start at
Visits that add to cart6.6%Practices 3 to 6: people aren't sure about the product or you
Checkouts that get finished (completed divided by reached checkout)47%Practices 1 and 2: something at the end puts them off
Orders per 100 visits, phones and computers1.5 on phones, 1.3 on computersPractices 7 and 8, if your phone number is well below your computer number: the store is harder to use on a phone
Visits (Shopify calls them sessions)Varies too much to compare. Under a few hundred a month, you can't judge the rest yetPractice 10: too few people are finding you

If everything looks about average, start at practice 1 anyway. Those visitors had nearly decided to buy.

1. Show the total and the delivery date before the cart

Don't panic about the size of this problem. Baymard estimates that about 70% of online carts get left behind. In its survey, 42% of US shoppers had left a cart simply because they were browsing. You'll never get that number near zero.

The reasons everyone else gave are worth your attention, though. Extra costs came first at 40%. Slow delivery was 20%, and 12% couldn't work out the total up front. All three have the same fix: tell people earlier.

Put shipping on the product page or build it into the price. Then swap "Ships in 3 to 5 business days" for a date someone can plan around, such as "Order today, get it by Thursday." Only promise what you can reliably deliver.

Test: On your phone, start on a product page. Count the taps until you can see what you'd actually pay, shipping included. If you have to type in an address to find out, this is your first job.

2. Keep checkout short

In Baymard's survey, 18% of shoppers had left because the site wanted them to create an account. Another 17% left because checkout was too long or complicated. Its testing says the ideal checkout has 7 or 8 boxes to fill in. The average US checkout shows about 15.

Three things to do:

Also look through your abandoned checkouts, where you can see what happened with each payment. In Baymard's survey, 10% of shoppers had left because a card was declined. So some of the people you think walked away may have been trying to pay you.

Test: Buy your cheapest product on your phone with a real card and count every box you have to fill in. Refund yourself afterward.

3. Answer every product question on the page

Your customer can't touch the thing or ask anyone about it. Nielsen Norman Group studied 49 online stores and sorted product page features into three groups. It's a handy way to decide where to spend your time:

GroupWhat's in itWhat to do
Must haveClear product name, images you can enlarge, price with any extra charges, options like size and color, availability, an obvious add-to-cart button, a short and useful descriptionFix these this week
Nice to haveReviews, extra images, video, zoom, related products, wishlistsAdd them as you can, reviews first
FancyVirtual try-on, 360° photos, customization tools, how-to videosOnly if they work flawlessly and your product truly needs them

Two more details are worth stealing. Show at least 3 to 5 images per product, and add video for your bestsellers and anything complicated or expensive. And get to the point in the first few lines, because people skim.

The questions people ask you are only half of it. The other half are the private doubts they never type: "This won't suit me." "It'll look cheap up close." "That's for people who know what they're doing." Good sales psychology brings that doubt into the open. Answer it right next to the buy button, with a measurement, a close-up or a line from a customer who thought the same thing.

Test: Read the last 20 questions people asked you before they ordered. Is each answer on the product page? Every one that isn't goes on your list.

4. Get five reviews on the products that matter

Northwestern University researchers found a product with five reviews was almost four times as likely to be bought as one with none. The boost faded quickly after that, and it was larger on higher-priced items. People were also most likely to buy at ratings of 4.0 to 4.7 stars, and less likely as ratings neared 5.0. So you can stop losing sleep over the occasional three-star review.

So get a few reviews on many products before you get many on one, and start with your most expensive. Show both the average and the number of ratings in your product lists. A 4.8 from three people and a 4.8 from 300 tell shoppers different things.

Ask by email once the product has had time to arrive and be used: "Did it do what you hoped? One honest sentence is plenty, good or bad." Don't buy fake reviews, write your own or pay people for positive ones. US rules ban all of that, including fakes written by AI.

Test: List your ten best-selling products by sales. Which have fewer than five reviews? Email their recent buyers today.

5. Prove there's a real business behind the store

In Baymard's survey, 13% of shoppers left over the returns policy. So write yours in plain language with a short summary. If you cover return shipping, say so outright. Something like this works: "Changed your mind? Send it back within 30 days. We pay the postage." In Shopify, you can edit your policies under Settings, then Policies.

Also add a footer link labelled "FAQ" or "FAQs," because that's where people look when they get stuck. Beyond that, the basics are your own domain instead of the myshopify.com one, a real About page and a contact email someone answers.

Test: Show your store to a friend for 30 seconds. Then take the phone away and ask: who runs this, and how would you get your money back? Blank looks are useful information.

6. Make products easy to find

If you sell 15 things, you can skim this one. With more products, it moves up the list.

Test: Open Searches with no results in the same Behavior reports. These are people telling you exactly what they wanted and didn't get. Fix the top five. Either connect the word to a product or note what you might stock next.

7. Buy from your own store on a phone

Phones bring in 66% of visits at the typical Shopify store, and you probably run yours from a laptop. That gap hides a lot.

Two habits cause trouble. Homepage slideshows that change by themselves send people to places they didn't mean to go, yet Baymard found them on 52% of the mobile sites it checked. And chat buttons that stay fixed on screen often cover important content on a phone, while chat boxes that pop up interrupt people. So link to chat from your header, footer or help pages.

Test: Find your top three entry pages in Sessions by landing page. Buy something from each using one thumb. Any time you have to pinch, scroll sideways or close something that's in the way, write it down.

8. Keep pages fast

You don't need to become a speed expert. Google's targets are simple. The main content should load within 2.5 seconds. The page should respond to a tap within a fifth of a second. And things shouldn't jump around while it loads. PageSpeed Insights will check any page for free. Test a product page, because that's where people decide.

If you're on Shopify, it shrinks your images automatically. Shopify says the biggest factors are your theme, your apps and any outside code you've added, so look there. The web performance reports show how changes like installing an app affect your store, though the data can lag by up to 36 hours.

Test: List your installed apps and ask of each: what number does this move? Remove the ones you can't answer for. Check that nothing has broken, and compare a couple of days later. It may also save you a monthly bill.

9. Make the store usable by everyone

Plenty of owners skip this because it sounds like legal homework. It's also a sales problem. Someone who can't read your pale gray text or hit your tiny button can't buy from you.

There's a legal side too. The EU's accessibility law covers online stores. Very small service businesses are exempt: under 10 people, with yearly sales or a balance sheet of no more than €2 million. If you sell into the EU and might be over that line, ask someone qualified.

The usual standard is called WCAG. Three checks cover a lot of ground:

Test: Put your mouse aside. Press Tab to get from a product page all the way to payment. If you get stuck or can't see where you are, so will someone who can't use a mouse.

10. Help Google, and AI assistants, read your products

A small store needs three things.

Product data that machines can read. This is called structured data. It lets your price, stock, ratings and shipping appear right in Google's listings. Google notes that a plugin for your platform is often the easy way to add it. To see what you have, paste a product address into the Rich Results Test.

Real links to every product. A "Load more" button works well for shoppers. But Google's robots don't click buttons, and they generally follow ordinary links. So keep the button if you like it, but make sure every product can also be reached through a normal link.

Accurate details. People are starting to ask AI assistants what to buy, though those visits are a small share for now. The preparation is the dull part you need anyway: correct prices and stock, and descriptions that answer specific questions. I wouldn't build anything special yet.

One related fix pays off quickly. Send ad clicks to the page the ad promised. Someone who tapped an ad for a blue raincoat doesn't want to start again from your homepage.

Test: Search Google for `site:yourstore.com "a product name"`. If your own product doesn't appear, start here.

11. Stay useful after they buy

These people already trust you, and you've already paid to find them.

Repeat orders and bigger orders are their own subject. That's when you stop fixing leaks and start asking how to increase ecommerce sales from the people you already have.

Test: Place an order and read everything your store sends you. Would you feel looked after?

12. Ask the people who are leaving

This is the only practice that tells you which of the other 11 matters most for your store. It sits last because you'll just hear "shipping" until you've fixed the obvious leaks.

Put one question on your best product page, shown after someone's been there a while:

What's the one thing holding you back?

Make the answers tappable: whether it'll suit me, the total cost, delivery time, whether I can trust this store, just looking. Keep it small and out of the way, so it doesn't interrupt people like those pop-up chat boxes. A survey app can do this. Once the answers start repeating, you'll know what to rewrite first.

A survey has one weakness: the person who told you they weren't sure still leaves unsure. If your shoppers need an answer right there, you need something that can reply. That's what we built Possible to do, and yes, it's our own product, so weigh this accordingly. It's an AI sales assistant for online stores built on real sales psychology. It asks each visitor the right question at the right moment, remembers the answers and guides them to a sale. It's free to start. If you sell three products and the choice is obvious, you won't need it. Finish the list above first.

What can wait

All of these are good advice for someone. If you're a small team, here's when each becomes worth your time:

PracticeWhen it's worth it
Virtual try-on and 360° photosWhen returns over fit or looks are a real cost, and you can make these work flawlessly
Building a customer communityWhen your product has hobbyists who'd talk to each other anyway
24/7 customer serviceWhen you get enough orders overnight to justify it. Until then, promise a reply within one working day and keep the promise
A loyalty programWhen you have a decent number of customers who've bought twice
New AI checkout toolsWhen your platform makes it a setting you can switch on. Get your product details right first
Testing two versions of everythingWhen you have far more orders than most small stores get in a month. Until then, make the obvious fix and watch your numbers
A redesignWhen the store looks broken. Otherwise the same doubts follow you to the new theme
Fancier customer account pagesWhen you have enough repeat customers logging in to notice

And one to skip for good: fake urgency. Low-stock notices are fine when they're true. A countdown timer that resets itself is a lie, and anyone who comes back tomorrow can see it.

If you only have this week

  1. Monday: Run the tap test from practice 1. Put shipping and a delivery date on your best product page.
  2. Tuesday: Count your checkout boxes, check your sign-in setting and buy from yourself on your phone.
  3. Wednesday: Read your last 20 questions from people who hadn't bought yet, and answer them on the page.
  4. Thursday: Fix the top five searches with no results. Email recent buyers of your most expensive product for reviews.
  5. Friday: Run a product page through PageSpeed Insights and remove an app you can't justify.

Don't leave this until the rush. Anyone working on black friday conversion should get these basics in place well before the crowds arrive. The busiest week of your year is a bad time to change your checkout. And if you're on Shopify and this list is done, the question becomes how to increase sales on shopify from a store that's now ready for it.

Questions about ecommerce best practices

What are the most important ecommerce best practices?

Showing the full cost early, keeping checkout short, answering product questions on the page, collecting real reviews and making sure it all works on a phone. In Baymard's survey, extra costs are the top reason people leave a cart, so that's where I'd begin.

Do these best practices apply to a small store?

Mostly, but be selective. A lot of this advice comes from testing large sites. So advice on filters, category levels and customer account pages assumes you sell a lot of products. A store with 15 products can skip most of that and spend the time on its product pages and checkout.

What is a good ecommerce conversion rate?

A conversion rate is the share of visits that end in an order. In Littledata's numbers, half of Shopify stores get fewer than 1.3 orders per 100 visits, and it takes 2.7 to reach the top fifth. What you sell changes this a lot. Typical fashion and home stores get about 1, while beauty and skincare stores get 2.6.

What are the 7 C's of e-commerce?

There's no single version. One agency's list is convenience, content, customization, customer care, community, communication and connection. It's a fair reminder of what a store should cover, but it won't tell you what to fix first. Your own numbers will.

What are the 7 pillars of e-commerce?

There's no agreed list, so don't worry about the label. Pick the practices your numbers point to.

Is eCommerce still worth it in 2026?

People are still buying. The US Census Bureau estimated that online sales in the second quarter of 2026 were up 12.2% on a year earlier, against 6.7% for all retail. Online sales made up 17.1% of the total. Whether it's worth it for you depends on what's left after product, shipping and ads. Work that out per order before you spend on anything.

What are the top 5 items to sell online?

Amazon says the categories known to sell well in its store include electronics, fashion, home and kitchen, beauty and personal care, and health and wellness. Those are also crowded, and big demand doesn't mean a small store can win. I'd look for something with healthy profit where you can give people a reason to buy from you and not from Amazon.

Go deeper on conversion tactics.

More guides for online stores

Get new guides by email

A practical conversion guide when there’s a new one. Unsubscribe anytime.

In this guide
  1. Where to start if you only have an hour
  2. 1. Show the total and the delivery date before the cart
  3. 2. Keep checkout short
  4. 3. Answer every product question on the page
  5. 4. Get five reviews on the products that matter
  6. 5. Prove there's a real business behind the store
  7. 6. Make products easy to find
  8. 7. Buy from your own store on a phone
  9. 8. Keep pages fast
  10. 9. Make the store usable by everyone
  11. 10. Help Google, and AI assistants, read your products
  12. 11. Stay useful after they buy
  13. 12. Ask the people who are leaving
  14. What can wait
  15. If you only have this week
  16. Questions about ecommerce best practices