There’s nothing that always works and pretty much nothing that never works either. Websites are highly contextual.

That being said, there are tests that tend to have a very high win rate. These are the test ideas that, while they don’t work 100% of the time, work more often than not.

Naturally, everything depends on the specific implementation — a good idea implemented poorly will not yield any results.

The following 20 testing ideas come from our own client-based research done over the years.

Note: For the full descriptions and examples for each idea, refer to the full article linked below. I’m including just the most interesting and useful stuff below.

Test idea #1: One static image with a single value proposition is better than an auto-rotating slider.

In general, don’t use auto-rotating sliders (aka carousels). Though there are undoubtedly examples of them working better than static images (rare as they are), for the vast majority of sites, they are a usability nightmare.

Implement a simple, static hero image with a strong value proposition. Here’s a good example from Acquainted.


Test idea #2: Proper value proposition beats no value proposition.

If you’re Amazon, you don’t need to explain yourself. Most ecommerce stores don’t have that luxury and have a large portion of visitors that have never heard of them before.

You have to present your value proposition as the first thing the visitors see on your homepage, but should be visible in all major entry points of the site (category and product pages).

What should your value propositions convey? Relevancy, quantifiable value and unique differentiation.

Here’s a good example from Podia.


Test idea #3: Prominent contact information

Although it seems like a small thing, putting your phone number and email address on the top of your site usually boosts conversions. It’s a trust thing.

Test idea #4: Prominent free shipping information

Charging for shipping is a conversion killer.

Sometimes, it might be near impossible to make free shipping profitable. However there are strategies you can experiment with:

  • Establish a baseline: Compare conversion with and without a free shipping offer.
  • Create thresholds: Increase the minimum order value required for free shipping, and test the improvement in margin.
  • Set restrictions: See what kind of improvement you’ll get by offering free shipping only on select products where it is profitable.
  • Enact price increases: Increase all your product prices to compensate for the loss you take on free shipping and see how your profit compares.

Test idea #5: Prominent section for sales and specials

Various studies have repeatedly found that about half the online buyers would only buy discounted products, except under exceptional circumstances. Around 60 percent say they are looking for a section that identifies sales and specials.

Ultimately, do what’s right for your brand, but it might be something worth experimenting with.

Test idea #6: Emphasize wide-appeal products

The goal of the homepage is to get people off the homepage. The best way to do it is to get them to click on an offer they’re interested in.

Look at your best-selling products or best deals, and emphasize them on the homepage above all other (make them #1 in the visual hierarchy).


Test idea #7: Make the search bar more prominent

If you sell products that people know to search for, making the search bar bigger and more prominent tends to work great. Think Amazon-style search bar — the center of it all.

Test idea #8: Add a site-wide benefits bar below the header

“Why buy here and not from Amazon?” is a question that all ecommerce sites must answer. A prominent benefits’ bar above the fold is one way to make your case.


Test idea #9: Level-up your product descriptions

Product descriptions matter. The role of product copy is to give buyers enough information, so they could convince themselves this is the right product for them. Clarity trumps persuasion.

If you sell stuff you don’t make, don’t just repeat the manufacturer’s canned descriptions. Add your personality and recommendations. Tell the customer why you personally recommend this product and how it will help them.

Test idea #10: Use product videos and test autoplay with captions

Images are good, but everything indicates that video is the future. Photos have their limitations, video is the next step before actually touching and feeling.

If you’re not doing product videos yet, do them for at least part of the inventory and see if it makes a difference.

Test idea #11: Ask for email first on checkout

This is so you’re able to do email retargeting if they abandon their cart.

Make sure the first follow-up email goes out ASAP. If they complete the purchase somewhere else, it’s over. If the purchase takes place, send one or two more follow-up emails (might include a coupon).

Track the effectiveness of those emails (open, clickthrough and conversion rates).

Test idea #12: Make the shopping cart persistent

People comparison shop. A common behavior is that they add products to cart on a site, so they can return to it later.

If upon their return, they discover the contents of the shopping cart has expired, they will not start from scratch (too much hassle).

Giving them the option to send the cart contents to email (later retrievable through a link) is a smart way of staying on the shopper’s mind.

Test idea #13: Create visually prominent and clear progress indicators during checkout

Are we there yet?

We like to know how much longer something is going to take. This is why numbered lists are better than unordered lists and why you should have clear progress indicators on your site.


Test idea #14: Clearly address purchase uncertainties

Is this safe? Can I return it? When will I get my stuff?

Make a list of the most common objections and doubts, and address them on the product pages and in the shopping cart.

Test idea #15: Expand payment options

Adding options like Paypal or Amazon Payments to credit card payments will help you win over some customers you’d lose otherwise.

Test idea #16: Cut form length

Make your forms shorter.

This is the very reason people prefer to buy from Amazon.

Their shipping and credit card information is already there, so they will save themselves from the hassle of completing forms. They’re even ready to pay a higher price just to save a couple of minutes.

Test idea #17: Don’t force people to register, add it in the backend

Don’t force people to register!

Instead, offer the option to register if they want, but create an account anyway for those who opt for guest checkout. They will enter their email and name anyway. You just have to generate a password and email it to them once they complete their order.

Test idea #18: Display reviews more prominently

People use reviews a lot. Even while they’re shopping in brick and mortar stores they read reviews online. You’re probably doing it too. I know I am.

Bottom line: Start gathering and showcasing reviews on your site.

If you sell commodity products and can’t get users to write many reviews, you may want to look into pulling reviews from an external site to have more of them.

Don’t delete negative reviews they actually help sales if there are only a few of them.

Test idea #19: Intelligently upsell

Upselling and cross-selling will boost your average order size.

Here’s the rule of upselling: You only offer related products (Apple offers smart cover for iPad and doesn’t try to sell you an iPod), and the offer must be at least 60 percent cheaper than the product they just added to their shopping cart.

So, if they’re buying pants, upsell a belt.

Brick and mortar supermarkets will try to upsell you while you’re checking out (grab a candy bar while you wait). Online, don’t do it. Focus on getting them to check out.

Idea #20: Clear, big calls to action

The user experience must be smooth. Smooth in the sense that they should never have to look for something. It should always be obvious how things work.

If people need to look for “add to cart” or “checkout” buttons, you’re failing miserably.

Those two are the most important buttons in your store. You want them big, bold and prominent. Avoid text links.

Check out how Patagonia has evolved their product page. Notice how the newer design on the left, and the red color that sticks out is the add to cart button and the cart itself.



Take a few ideas and test. Don’t copy blindly.

Then take a few more.

Learn what works for you and build on that.


Top 20 A/B Ecommerce Test Ideas

PS. And please don’t do any of the above if you haven’t done any research. Everything should be based on research. Only when you know what is not working can you start fixing it.