Internal graphics, banners and ads for promotions and site features are fairly common on e-commerce sites and can – when linking to relevant content – be immensely helpful to users. However, during our latest e-commerce usability study, a large number of the test subjects were led astray by poorly placed and designed internal ads. In some cases this even resulted in site abandonments.
We found the following 5 types of internal ads and banner graphics to cause particularly severe usability and user experience issues during testing of 19 leading e-commerce sites.
Note that the use of “ads” throughout this article refers to any type of banner or graphic. By “internal” ad we refer to any such banner or graphic that publicizes a feature or product of the site instead of an external 3rd-party offering.
When internal ads are placed above a product list, they are likely to introduce scope confusion and lead users to believe that the ad is somehow related to the product list. Any ad text may be misinterpreted as a header for the product list, and buttons and links may be mistaken for filtering and sorting options, as the following examples illustrate.
It’s easy to see how a hurried user can mistake such banners as related to the product list, and it’s therefore recommended to avoid placing ads above product lists altogether – neither on search result pages nor on category pages. The internal ads depicted above are fairly relevant to the user’s context and their presence on the page is thus justified – it’s not that they shouldn’t be on the page, they just shouldn’t be placed above the product list.
When ads are placed within a product list, users tend to interpret it as the end of the list. In the user’s view, the ad represents the beginning of a new page element and the ad therefore also comes to signify the end of the current page element (i.e. the product list). This problem is only exacerbated on devices with fading scrollbars, as the one visual clue indicating that plenty of additional content is available on the page will often be hidden.
Users generally perceive lists of product and categories as a “collection” and expect it to be represented as a whole. Therefore, when subdividing such lists by intersecting them with internal ads, a good deal of users will perceive it not as a singular list divided by an ad but rather as one list, an ad and then another list. As hinted, this issue isn’t specifically related to product lists but also applies to custom category pages and help sections, causing users to miss out on important navigational content and critical site assistance.
We furthermore observed that when ads are placed below a product list, users are more likely to mistake any text links in the ad for sub-categories or filters for the product list.
Users seem to be more likely to perceive links placed below the product list (and above, as described earlier) for filters and sub-categories. However, graphical ads placed at the end of a product list didn’t suffer from this, but were instead perceived to signify the end of the list, as described earlier.
Graphical ads may therefore purposely be placed below product lists to signify their end – just be sure to avoid text links so the ad isn’t confused for sub-categories or filtering options. In other words, it’s OK to place ads below the product list as long as they are properly designed (i.e. graphical in nature).
When benchmarking 50 major e-commerce sites, we found 43% of them to violate the principles described in the this and the above sections by either placing graphical banner ads above or within the product list, or by locating text ads below it (see all 52 product list examples). So be sure to double-check your e-commerce design for these three pitfalls in particular.
Some of the test sites had overlay dialogs (or “lightboxes”) appear upon page load – often containing surveys, contests, newsletters, promotions, country-selection, etc. This didn’t just happen on the homepage but also on category pages and product lists as well. Each time it happened, the vast majority of the test subjects would reflexively close the overlay dialog, not even glancing over its contents.
What’s more, the subjects would get annoyed with the site for bothering them, even though they – as just mentioned – never actually read what they were being bothered with. Some subjects went as far as referring to any overlay dialogs as “spam,” which gives a pretty good idea of just how much users loathe these dialogs and how poorly it can reflect on sites that use them.
It’s therefore recommended to avoid displaying overlay dialogs on load altogether, especially if the content is highly relevant to the user (again, most users will ignore it and just close the dialog instantly thus missing out on important content). And of course if the content isn’t of great importance, should you really be bothering the user with it?
It shouldn’t come as a surprise that overlay dialogs aren’t popular with users, as they are essentially the in-page equivalent of the much hated popup-windows that were rampant on spammy sites during the 1990s and early 2000s. Surveys, newsletter sign-ups, contests, discounts, you name it – most users will close the overlay dialog without even realizing what it’s about.
It should be noted that overlay dialogs may be used to display content upon user request – as long as the overlay dialog doesn’t appear upon page load it is absolutely fine. So by all means do display inline help, cart contents, etc. in an overlay dialog, as long as it happens as a response to a user action (e.g. mouse click) and not upon page load.
Of the 50 benchmarked sites, a hefty 40% flash an overlay dialog in the user’s face when they visit the homepage or a category page. Unfortunately, users ignore this younger cousin of the ’90s popup-window, reflexively closing these overlay dialogs despite their often perfectly relevant content.
“Wha– Hey… that’s some pretty aggressive ads they show here,” a test subject said as Drugstore’s homepage loaded, “Argh, this distracts me.” Internal ads and banners were generally perceived negatively by the test subjects, which confirms similar observations made during both our E-Commerce Checkout and Mobile E-Commerce studies. However, a new addition to these observations is a particularly marked distaste for ads on the homepage specifically.
When an ad was placed in a prime content location on the homepage – typically in the middle column, and particularly in the upper part of the page – the subjects would often comment negatively upon it. Since the homepage is many users’ entrance page such ads can therefore set a poor first impression of the site, potentially rippling throughout the user’s entire shopping experience.
Ads on the homepage can in other words set a negative tone for the entire site – a poor first impression that permeates through the rest of the user’s browsing experience. What’s tricky here is when something is perceived as an ad vs just a nice graphical representation of a product, category or site feature.
Users don’t make the technical distinction between internal and external ads, but rather respond with a gut reaction that has been conditioned over years use of websites with banner graphics. It therefore doesn’t matter if the content is relevant or appropriate – if it looks like an ad, it will generally be perceived as one.
We found the subjects more likely to perceive graphics with headers boasting about site-wide competitions and discounts as ads, whereas graphics showcasing specific products, categories and seasonal campaigns, were generally perceived as “legitimate” content. However, there were multiple exceptions and we therefore strongly advise designers to be very careful when designing their homepage and its graphics so that they don’t come across as ads.
In general, the primary header of the ad shouldn’t be overly boastful. Font choices and background graphics also greatly influence what is and isn’t perceived as an ad. If a graphic largely adopts the visual characteristics of the site (background, text colors, font choices, etc) and avoids visual borders or similar visual boxing, it’s less likely to be perceived as an ad. Notice how the prominent smartphone offer in the above BestBuy example adopt the general visual characteristic of the site and thus feel like a more natural part of the page than the “Deal of the Day” and “Geek Squad Tech Support” boxes lower down the page.
40% of the benchmarked e-commerce sites unfortunately fall victim to the allure of having highly graphical ad blocks in prime content locations on their homepage (see all 51 homepage examples). While numerous of the other benchmarked sites also had graphical blocks on their homepage, they had designed them much more carefully so they’re unlikely to be perceived as ads but rather as legitimate and helpful content.
In an effort to solve the logistical issue of having many different departments frequently in need of promoting offers or site features, it can be tempting to carve out blocks of fixed position and size in various places on the site, and then allocate these spaces for internal “ads” and campaign graphics. This is convenient for everyone: design just has to carve out a block on the page, development doesn’t have to deal with trivial site updates, and marketing can do frequent site updates and upload graphics as they please.
However, internal ads spread across the site this way are likely to cause issues since it’s difficult to foresee all the contextual issues they may provoke (see earlier Pottery Barn and Toys’R’Us examples). These “ad blocks” can also turn out to produce misleading design clues (see earlier Newegg and Pixmania examples). Finally, if highly graphical ads boasting about discounts and competitions are inserted, the page may appear cluttered and overly salesy (see Drugstore and BlueNile examples), which is likely to set a poor first impression of the site.
Therefore, if opting for this strategy, it’s particularly important to avoid the don’ts presented throughout this article. We can summarize these don’ts of internal ad design as follows:
Luckily, when great care is taken, internal ads can be implemented successfully without stirring up negative emotions or misleading the user. (Indeed, as we’ve seen with the BestBuy example, carefully designed internal ads can be perceived as useful content.)
So what is good internal ad design? As already mentioned, carefully designed graphical banners publicizing specific products, categories and seasonal campaigns are perfectly fine, even on the homepage (in fact, large bespoke product imagery was very well-received by the subjects). Also, graphical ads placed below product lists are perfectly fine and can help signify the end of the list. In addition to this, we observed that internal ads placed in a sidebar didn’t cause any issues.
Ads placed in a sidebar at either side of the product list were never interpreted as anything but ads. In other words, you’re free to place any type of ads in sidebars, as long as they aren’t intermingled with filtering tools or sub-categories, and as long as it’s not the only path to important navigational features (due to the risk of banner blindness).
It’s recommended that the ad remains somewhat graphical if placed immediately above or below categories or filters (to help differentiate the two), but during the study the subjects were generally quite good at making the distinction, as long as the ad was in a sidebar and not above or below the product list.
The do’s of ad design can thus be summarized as follows:
Join 24,000+ readers and get Baymard’s research articles by RSS feed or
Topics include user experience, web design, and e-commerce
Articles are always delivered ad-free and in their full length
1-click unsubscribe at any time
Thank you very much for the article, it’s a well-written piece. I had a few questions:
First, the short one:
I see that most of your examples seem to come from desktop experiences. Given the growing importance of mobile web, do you believe your research still applies? What are the ‘safe zones’ for internal marketing for a web app in a single column, for example?
Second, a more involved question:
I’m an interface designer and am currently facing an issue at work where the marketing team is absolutely convinced that we must use a full-page overlay dialog on page-load of our web app to advertise the existence of our iOS app (download and app use is currently a company priority). To their credit, they acknowledge the unpleasantness of the page-load overlay and are specifying that it should be shown only once for a given IP address, and only to users accessing our web app on an iOS device (never on desktop, or non-iOS mobile devices – only where conversion is possible). They understand that there will be an increase in bounce rate that results from this design decision, but are of the opinion that conversion is so important that it’s a cost worth paying for the extra organic downloads it could generate.
I absolutely hate full-screen on-page-load overlays as a consumer, I think of them as distasteful as a designer, but I do understand the marketing team’s reasoning even if I tend to disagree with the ‘acceptable cost’ bit. Any advice on how to proceed in this particular situation? Any data to add an edge to my argument (or to reassure me of theirs) would be most appreciated. I’m going to check out your benchmarks and reports now to see if I can spot anything I can bring back.
Fantastic job on a very useful resource – thanks again!
Yes they still apply to the full extend for the desktop design. When we conducted our large scale mobile usability study (http://baymard.com/mcommerce-usability) we found that ads on mobile e-commerce sites generally perform even as user’s don’t have nearly the same degree of page overview, and often get “lost on the page” – making it even more likely that internal graphics are misunderstood as the primary in-page content.
In regards to the “download app” splash pages the test subjects encountered several of those during our mobile study (see the M-Commerce Usability report, guideline #28). As you mention displaying them only once is key. Two crucial elements in determining if the “download app” is worth promoting as heavily as you mention, are:
1) exactly what benefits the user will get from downloading the app? (just another layout or? + remember to summarize the benefits at the overlay/ splash page))
2) It also depends on the site-context, is your site/content something that users will return to to on a regular basis, or is it only once in a while. Most test subjects expressed that they expected the app version to be faster and would therefor consider it for sites which they’d use on a regular basis (for most of them this would be defined at least twice a week, i.e. most e-commerce sites wouldn’t qualify).
© 2021 Baymard Institute US: +1 (315) 216-7151 EU: +45 3696 9567 firstname.lastname@example.org