News Feed Design Experiments: February 2017

As you know, we’re heavily investigating and building out a new way of showcasing actionable data in a friendly and familiar form: The glorious news feed.

We started experimenting with a newsfeed-ish prototype late last year and provided a simple retrospective of that work. We’ve continued to iterate on our growing data models and all of the data that we’ve ingested into our system(s) and wanted to share our progress as we put it together.

Again, this is super early stage prototyping so nothing is locked in stone here but we are very excited about our early experiments and continue to riff on it every week.

If you want to read some of our previous thoughts on this, check out these posts for context:

  1. On Developers and Social Networks [9/9/16]
  2. Thoughts on a Developer-centric News Feed [9/28/16]
  3. Retrospective: Building a Feed [11/1/16]
  4. It’s Time for a Technical Newsfeed [1/20/17]

As you can see, we’ve been thinking about this for months and have begun to build out these thoughts and concepts into something usable. I’ll provide a little bit of explanation but not much, at least for now. Consider this just a visual teaser of much more to come!

But, first, of course, a gif:

So data, much wow!

Above and below exclude the greater context of the UI, but, you’ll get the picture.

Card: Commit detail

What we’re exploring is how to showcase data that you want and need into “cards,” as we’re currently calling them. One of the most important questions that we’re trying to answer is what specifically we want to show so that the end result is actionable insights, which is part-and-parcel of engineering ops.

Card: Even more detail

Is more data in each card better? We won’t know until we have our internal tests running over a good period of time as well as our early testers as we gather their feedback.

A few more views:

Card: Details with GitHub data, views

Card: Details with comments

Again, we’re quickly moving into these implementations with a very open mind as we’re not entirely sure, yet, what’s going to work the best. Experimenting with both the data that we show and how we create visual cues and visual signals for action is of utmost importance:

Cards: Experimenting with visual signal(s)

We’ll keep you updated with our work and we’d love any visceral feedback that you have to offer! Very excited to be iterating on something that will provide an incredible amount of value in a way that’s relevant, familiar, and ultimately super-practical and very useful.