Curbed on Chorus, the start of a new era at Vox Media

On Monday, we relaunched Curbed on Chorus, the final brand from November 2013's Curbed Network acquisition to migrate to Vox Media's platform. Though today's launch might look like any other Chorus site on the surface, it marks a significant evolution of how we launch brands at Vox Media on our new, unified version of Chorus.

To explain, let's go back to March of last year, when we kicked off a project called Unison.

What is the Unison Project?

We have traditionally done a lot of custom engineering work for each brand that joins Vox Media. Starting with the launch of The Verge in 2011, we've focused on creating brands that are highly individualized, both in terms of visual design and site functionality. Our brands are powered by Chorus’s common set of publishing and community tools, but most of the presentation and user-facing functionality has been custom-built specifically for each brand.

This approach worked when our focus was on improving our individual sites, but has made it hard (read: very inefficient, near impossible) to consistently iterate on our underlying platform. If Racked wanted a new photo gallery feature, for example, our two options were either (1) only make updates for Racked, further deviating it from the rest of our sites or (2) spend a very long time redoing the same work seven times over, each in slightly different ways. This was clearly not sustainable.

We needed to invest in making Chorus a platform on which media brands can be more efficiently designed, built, launched, maintained, and evolved. The level of effort to release or iterate on a product feature or ad product must not increase with the number of brands running on Chorus. We needed to shift the creative focus of our design and engineering teams away from maintaining seven discrete brands and toward building systems that enable numerous, highly-individualized brands to be created and operated without enormous human effort.

So we built a more unified, highly-customizable Chorus, with a flexible set of design components that can scale for many different types of visual designs. With the click of a button, we can add a new Chorus site. We now have one unified set of publishing, community and advertising tools, contained within one singular, shared code base.

Goals

The project to unify our platform had two guiding principles: scale and speed. We wanted to more efficiently add brands to Chorus while at the same time increase the speed of iteration on the underlying platform for all brands running on Chorus.

In addition to these guiding principles, some of our other goals:

• Mobile-first design: Everything about the Unison framework design was done from mobile-out. Because 50 percent of our traffic now comes from mobile, providing a great reading experience for users on mobile devices was fundamental. Much of our early exploration was centered around optimizing the mobile article.
• Simple, reusable, baseline design and functionality: Since a goal of the Unison Project was to create a framework that we could continue to evolve and improve with data, we kept our baseline simple. We focused on creating reusable components that support a wide range of content types. Much of this work started with identifying what's currently working well across several of our sites, while also making subtle refinements.
• Finding key elements of brand identity: While creating a scalable framework built from reusable components was a core goal, we also needed to identify areas for our brands to differentiate themselves within this system. We ultimately need to support at least eight brands on this framework, and potentially many more in the future. Throughout our work on this framework, we consistently identified scalable opportunities for our brands to express themselves. This includes variations in layout and opportunities to be expressive with typography, shape, and color. This system will evolve over time as more brands are added to the new framework.
• Performance: Performance is a company-wide goal at Vox Media and the Unison Project operated with a strict performance budget and made sure to deliver our web pages as quickly as possible. And we delivered! Compared to the pre-Chorus Curbed, the new site is loading 50 percent faster and seeing a decrease in page size between 60-70 percent. (Spoilers: We'll have a blog post from our performance team droppin' later this week that will detail the Unison performance optimizations).

Designing brands, not websites

So how did all this change the way we approached Curbed's design? The old model for launching a brand was all about building a website. With the Curbed relaunch, we were able to focus our design energy instead on designing the brand overall. Our design process wasn't about solution design and reinventing article pages or topic pages; it became about an expression of the brand on an already-solid framework. (Related: Behind the scenes of Curbed's design process)

Our initial design comps looked at how the brand would live in places like Apple News, Snapchat Discover, Instagram/Twitter/Facebook, even on tote bags. It wasn't until the end of the process that we applied the design direction to the Chorus framework. At that point, we were able to evaluate what was missing in the core platform's design system, and prioritize which configuration variables to add.

Because of the platform-focused approach to the design system, our brand designers were able to directly dive into the Chorus design configuration tool and make decisions without needing to file an issue for engineers for each little tweak.

There's also the added benefit that every bug we fix for this launch will impact all future launches. Meaning, in theory, each launch will get simpler. Though we do realize there's complexity in every future change we make, and every change is now higher risk. A worthwhile risk, in our opinion.

What's next?

Now that we have a singular version of Chorus, and have launched a brand on it, it's much easier to launch new brands from the technical and design perspective. We can iterate more on our platform across the board, and universally benefit all our brands with each new iteration.

Over the next year, expect to see more brands moving over to the new and improved version of our platform. We'll approach this migration process iteratively, too, so you won't see it all at once, but you should notice zippier performance over time.

Oh, and if this sounds interesting to you, we're hiring.