In my daily trips around the web, I've noticed that many of the most popular ecommerce sites (e.g. Walmart, Target, Amazon, etc.) are removing CTAs (e.g. "add to cart") from search results -- and I can't help but wonder why.

Is there any existing research that indicates improved click-to-sale conversion rate by removing the "Add to cart" button from product listings?

As UX designers, have any of you made this change on your site? What was the perceived problem that drove the change? Did it work?

  • Is there any existing research that indicates improved click-to-sale conversion rate by removing the "Add to cart" button from product listings?
    – UXeMpath
    Commented Apr 5, 2016 at 18:35
  • It's the result of continuous A/B and multi-variate testing. It doesn't matter if some people don't like it, the version that makes more money wins. Commented Apr 5, 2016 at 18:45
  • The nature and phrasing of this question will likely result in it getting closed for being too opinion-based, leading to a discussion based on subjective feelings, which is not what this site's for - ux.stackexchange.com/help/dont-ask You should consider reconstructing the question so that it can be answered objectively in a way that provides general and lasting value for future readers ux.stackexchange.com/help/on-topic
    – dennislees
    Commented Apr 5, 2016 at 18:50
  • As consumers and online shoppers, I think we're all experts of a sort. I'm interested in knowing why one would prefer/not prefer the CTA in product listings and why. I think this could help answer a design question and help with a solution. Once armed with a bit of info, A/B or multivariate testing might be more valuable.
    – UXeMpath
    Commented Apr 5, 2016 at 19:09
  • I did test the addition of cart buttons on a category/browse page. Much to my surprise (and delight), it lost resoundingly. After several follow up tests and interviews evaluating other aspects of the category page, we came to the conclusion that it was too much to consider. Users were in the mode of evaluating and comparing at a high level and didn't want to think about a dozen add to cart buttons. Commented Apr 6, 2016 at 6:26

3 Answers 3


First thing first

Nobody is removing CTAs. In any case, some sites are (well, I think they did it some time ago) removing CERTAIN CTAs. In the specific case you mention, the Add to Cart button.

As to why are they doing it, it may be many reasons. But answering a part of your question

As UX designers, have any of you made this change on your site? What was the perceived problem that drove the change? Did it work?

yes, we did it. As a matter of fact, unless we're specifically requested by a client to include an Add to Cart button, we don't add it by default. Not only in search result, but also in default listings. We always recommend clients to NEVER include a direct buy CTA in any kind of listing (with some very specific exceptions based on testing).

The reason why WE don't recommend it:

  • Accidental add to cart action
  • Without proper information, clients may buy the wrong product, increasing returns and adding more costs and friction to the process.
  • Loss of proper selection of colors, sizes, types and more
  • Loss of ability to add packages
  • Loss or related products upsales
  • Clients can't read TOS, limitations and other legal specifications

just to name a few!

As you can see, what you may perceive as annoying in fact saves you and the site lots of problems and money. Furthermore, based on our own testing (which may or may not have similar results to the tests those companies ran), customer degree of satisfaction increased since there's more transparency and less mistakes, which in turn means more revenue.

In short

Again, I can't tell why those companies did it, I just can tell about the part I know. And based on our research, the easy answer to why they're doing it is... because it makes more sense. The strange part is why are there companies still having buy actions on listings!


As a general design question, I think it is not unusual for design trends to change, if only to adjust to the change in user behaviour since the last wave of design trends. So I suppose in general you can look at the 'trend' as being something that is part of a constant change in design ideas, and the other part being driven by specific factors in the human/web interaction.

In terms of the specific design trend you are describing, I think it also needs to be broken down into a few more components. Firstly you have to look at the differences between large e-commerce sites that aggregate lots of product brands and ranges like Amazon compared to the online stores of individual brands. Secondly you need to look at the impact on the existing users of say Amazon versus new users.

Assuming that you can get data on all these types of users, then perhaps you can provide a more objective answer of whether there is an overall 'trend', and what the impact to the users will be.

However, I think the general result of such analysis will show that if the users really want to buy the product, then they will find a way to do so. And if they can do something one way, it is not conceivable that they will still change their mind about it over time. As usual, testing on your own users is better than making inferences from general trends.


Exposure = Revenue (the business side)

The cognitive goods matrix:

A 2 by 2 matrix, with need/want on one axis, aware/unaware on the other. Exposure moves unaware to aware

  • There are goods you are aware you need. This is your shopping list.
  • There are goods you are unaware you need. This is the stuff you forgot to put on your shopping list.
  • There are goods you are aware you want, but you don't exactly need them. Like a book you have on your wishlist, but you have 3 other books in your shelf to read anyhow, so no urgency ordering.
  • There are good you are unaware you want, like anything on sale or chocolate truffles.

Exposure automatically promotes the unaware to aware status. Exposure can also lead you to buy things you don't necessary need ("Minimum order is £20, so I might just as well tuck this book I want in there").

Time on site = exposure

There is a maxim in commerce that goes: Exposure equals revenue. One way to increase exposure is to increase time out site. You don't want people buying from the search results page and checking out.

People spend more in a shop than they do online. This is simply because in a shop they are visually exposed to many more products than they are on a computer screen, and the physical nature of a shop means more browsing no search boxes. Financial advisors are unanimous about online shopping saving you money compared to a shop visit. Supermarkets constantly change the location of products so frequent visitors search harder, by that being exposed to more products.

How many times you entered a shop with a friend who was looking for something and ended up buying something yourself, despite this was never on the day's agenda?

Search result pages offer little opportunity for the business to expose you to more products - let it be related, those you previously searched for, or just random ones; when clicking on alternative products you are exposed to even more products. A product details page has much more space to do so, and users are likely to spend more time there.

The user side

I recommend reading chapter 2 in Designing the Search Experience, where a various search models are covered.

They all suggest that the search result page is not where people are likely to buy from. Then, there's the social effect - reviews are highly affective. Even if you are in an exact hit mode, you are more likely to use the previous orders page than the search feature.

So not having the button there is not exactly a usability tragedy from a user perspective.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.