To Pop or Not to Pop

The pop-up window had several significant limitations. The most significant of these was the nature of the pop-up window itself. Several years ago, as the web was still evolving, the presence of a pop-up window seemed to indicate the presence of important information that the user should attend to immediately. This notion probably stemmed from the familiar "alert boxes" found in both the Windows and MacIntosh Graphical User Interfaces (GUI's). However, the concept of the pop-up window was quickly co-opted by marketers as a vehicle for delivering advertisements.

Numerous research studies have supported the claim that users routinely expect that pop-up windows will contain advertisements. Many of these users will close these pop-up windows before allowing them to load. One member of the senior management team even admitted to closing pop-ups before they loaded, but stated that he let the pop-up cart load because he knew it was a shopping cart. We quickly pointed out the fact that users unfamiliar with the system would have no way of knowing that this pop-up contained shopping cart info.

When the pop-up cart was first conceptualized, one of the advantages put forth was its small size. It was believed it would load more quickly than a full page cart and be a better experience for the user. However, as new features were added, screen real estate in the pop-up cart window became extremely limited. There was talk of making the pop-up window larger, but this seemed to defeat the stated purpose and created new problems (such as obscuring the product page).

Since the regular web site pages were templated, navigational and design elements were reused on every page. Presumably, the user's browser would cache these elements and they would load nearly instantaneously after the first instance. In our preliminary tests of a full page cart versus a pop-up cart, we found that the processing time needed to launch a new window (and load its contents) was virtually identical to loading the same contents along with previously cached navigational and design elements. At this point, we eliminated the "speed of loading" argument as an advantage for the pop-up cart system.

Our next step was to examine elements of the pop-up cart system that we wanted to retain and/or enhance. The primary feature we addressed was the ability to calculate shipping charges from the shopping cart system (as opposed to waiting until the user went to the checkout process).

Figure 2 - Shipping Cost Calculation for Full Page Shopping Cart

One of the principle advantages of the original pop-up cart was to allow users to get information on estimated shipping charges prior to proceeding to checkout. This is an extremely important concept with regard to usability and cart abandonment. The most frequent point at which carts are abandoned is on the page where shipping charges are first revealed. Far too many sites make the mistake of making this disclosure deep within the checkout process. Our model allowed for shipping charge disclosure before even approaching the checkout.

This feature was so important that we believed it should be enhanced and its usability emphasized. With the pop-up model, the user had to locate a very small link indicating that they could click to add their zip code to determine shipping cost estimates. This link loaded another page within the same pop-up window where the user was instructed to enter a zip code (International users selected a country and postal code where applicable). Submitting this form then reloaded the original cart page with shipping charges now dynamically displayed in the order table.

It was our belief that we could increase the utility and usability of this feature by eliminating a step and allowing the user to enter a zip code directly from the cart to display dynamic shipping cost estimates. This method is more intuitively obvious and thus was hypothesized to be more usable by the average user.