Alastair Jardine, one of the architects of the Guardian of the user experience, speaking through our new video page design

We have launched a new video page last week here in the Guardian, which is the result of hard work by the core team, working on a guardian related functionality. co.uk. We had a number of design objectives and user experience, which we will explore in this post.

One of the main goals was to make the video more important. It seems obvious ("Video is a page"), but in the previous model, the first thing that presents the user was standfirst - a paragraph of text describing the video content. Whereas standfirst is good for SEO and is generally useful, this text is not as convincing as the video itself. He also had the unfortunate side effect of pushing the video at the bottom of the page, forcing the user to scroll to see the video really. The video player is relatively small - 480px wide - something that frustrates us, and not allow our excellent content to shine. now is much greater.

The final design puts the top of the video player larger and the left, and a new video carousel on the right. The carousel automatically displays videos of the same series as video and videos of automatically generated if the series is not. We rationalized that if a user visits the game content, who prefer to participate in the videos in the series and not just related videos. In this project, we also implemented a endslate "', which shows most popular videos on the time to look.

Many of these decisions were informed by data, either from our analysis, or by A / B testing We found a series of A / B testing to see what we can learn from the old design when we designed the new page. A simple and obvious way is to put the below video: we scored an extra 15% clicks on videos. We also learned that the recent default tab (as opposed to shareholders) resulted in a 44% decrease in the percentage of clicks.

Needless to say that these results hold in view of the guard, so that these behaviors are contextual and specific use case, and will be reviewed in the light of the new design. We keep a close eye on the performance and testing often generate lower earnings in clicks and interactions. Individually, these achievements do not make much (maybe 1-2%), but taken cumulatively, they do.
Lessons Learned ? The things we could have done differently? I want to make the process faster, more frequent deliveries with iteratively working towards a "final" design (in design ever done?). He also acknowledged that the two models have been appropriate for our content:. A more attractive for the publication of high-quality video and the other for viral videos, the size is not always important, but where is the path of discovery and content