When in Amazon AR View, the user is generally very clued into the system status and is kept informed about what is going on. The first thing that entering the view greets you with is a popup telling you to scan the floor, letting the user know that the feature needs user action, only to disappear if the user correctly completes the action. Once the floor is correctly scanned, there is a dot matrix overlaid on the floor to show the user where the item is to be placed, with a very eye-catching bright yellow outline of the product size with the text “tap to place”, making it clear to the user that the item is ready to be placed. Once the item is placed, and the user taps on it to enter editing mode, the item is outlined with a blue 3D box. This is especially important when there are multiple AR items placed, helping to tell the user which item is being manipulated. When moving the item with one finger there are straight arrows in each direction, showing the user how the item can be moved, and if the user adds a second touch input the arrows turn curved, immediately providing the input to show the user that they are to use two fingers to rotate the item.

The feature also does a good job at using language and concepts that are intuitive and familiar to the user. One big example of this is that even though the feature is officially called “AR View”, the term AR doesn’t appear anywhere while initiating or interacting with the feature. This is helpful as not everyone would know what that term means. Instead, the feature is dubbed “View in your room”, which uses terms more familiar to your average user. The use of radial straight arrows or curved arrows surrounding the item depending on how it is currently being manipulated matches with our cultural understanding of what different arrows mean. When there isn’t enough light to properly scan the floor, the error message that pops up includes a symbol of the sun, which the user would likely immediately relate to bright light. This match makes the popup error more instantly effective. 

When using AR View, there are not many sequences of actions that aren’t easily reversible. If the user entered the mode accidentally, there is a simple back button in the top left to exit that simply requires user confirmation to leave. If the rescan button in the top right is touched accidentally, you can easily back out by clicking “cancel” in the confirmation popup. If a secondary item is placed and the user wants to delete it, the process is simple as one needs to intuitively click the garbage icon once to initiate deletion and again to confirm.

AR View does an excellent job at conforming to industry standards in the symbols it uses. Every single one is a symbol a user would have encountered previously, and the usage aligns with what they would expect. There is a plus symbol to add more furniture items to the view, a sun symbol to support the light level error message, and the rescan symbol is the circular arrow commonly used to indicate reloading or redoing. A garbage can symbol is used for deleting items, and the back symbol is the very common left pointing chevron.

Users are always checked with to confirm they want to proceed with any action that would be inconvenient to recover from. A popup is utilized to confirm when the user initiates leaving AR View or rescanning the room. Similarly, confirmation is required to delete a placed item. The garbage symbol tile expands upon the first click to reveal the text “Tap to remove”. All of these steps go towards preventing user slips.

Reducing memory load is always a priority in UX and UI design. Amazon AR View generally does a good job of this by making actions and items visible to the user. There are not really any clever tricks I could identify that go towards reducing memory load; the AR View feature is inherently light on memory load.

One thing I appreciate is that there are two ways to enter AR View. The first is to simply click the “View in your room” button below the product pictures, but the second is on the second product picture. With a description of what AR View does, it prompts the user to try out the feature with “Does it fit? Use your camera to view in your room”. This flexibility in accessing the feature is definitely beneficial. With regards to this heuristic, my one critique is that even if an item of furniture is “hearted”, it does not take any priority in the list of items when adding a second piece of furniture. This decreases interaction efficiency as they need to hunt for the item they liked. “Hearted” furniture items should absolutely be prioritized in the list or otherwise differentiated from the other products.

The interface is very clean, without any distracting elements. The only things visible to interact with compose the core functionality of AR View, without any distracting decoration or unnecessary text. When not in product edit mode, the only two UI elements are the back button and the rescan symbol. I honestly think they go a bit too far in making their minimalist design, as they go so far as to remove the product brand and titles from the menu that lets you add additional items to the view. While there is certainly reasoning behind this design decision, such as an inability to fit the product titles into the current product display design, I think the solution should be to rework how products are displayed rather than remove the brand and product titles to make it fit into a more “aesthetic” layout.

Error messages are clearly expressed and described, with prompts to the user with the action that they should take to resolve the issue. When scanning the floor, if the camera is being moved too fast, there is a popup saying “Too fast. Slowly move the camera”. When the camera view is too dark, there is a popup saying “Too dark. Try adding more light”. If the user tries to remove the starting object, which AR View does not support, text pops up explaining that the initial object cannot be removed, which makes it immediately clear to the user which actions they can and cannot take.

When first entering AR View, the users are nicely instructed exactly what to do in order to set up the experience, with text and animation describing precisely what to do. If you persist and can’t correctly scan the floor, it prompts you again with a more detailed description of what you need to do: “Move your camera. Point camera at floor area with texture, patterns, or edges”. Upon the placement of an object, text briefly pops up saying “Drag and rotate. Move the product around the room” which helps users discover the manipulations they can perform on the placed item.