Enable optional spatial screen reader speech and directional navigation. Since participants found value from exploring websites with WebNExt, we recommend that screen readers consider features for spatializing all existing screen reader audio output, even when navigating via the current paradigm of an DOM tree-based approach. We also recommend that screen readers enable directional (up, down, left, right) navigation keyboard commands.
Enable user-driven customization. We recommend that future solutions allow users to customize audio cues for how web elements' 2D positions are conveyed. Some users may prefer using volume change or pitch to convey information, and may want spatial positions to be exaggerated more.
Build in additional feedback around visual concepts such as scrolling. Navigating visual content spatially is a very different approach than navigating via the DOM tree. Screen readers should provide feedback for users to know about visual concepts such as page scrolling and reaching the visual edge of the web page, as well as for other animations and visual effects.
Make spatial browsing toggle-able at input and output layers. If spatial browsing is offered natively as part of existing desktop screen readers, it should be an option that blind and low vision users can toggle on or off. Furthermore, enabling separately controllable spatial input and output allows users to choose their desired combination of input and output techniques.