Skip to content
Sam Hankins

Vox Media / The Verge / 2022

Make It Feel Like Posting

Designing a short-form publishing surface inside Chorus so The Verge could bring fast, voice-driven updates onto its own homepage without losing the structure of a real CMS.

I designed Quick Posts, a purpose-built composer inside Chorus, Vox Media's shared CMS for brands including The Verge, Vox.com, Polygon, SB Nation, Eater, and New York Magazine. The work was not to make a lighter article editor. It was to create a publishing surface that helped writers post quickly while the platform still handled the parts readers never see.

Composite image of The Verge relaunch homepage, Quick Posts cards, and the Chorus Quick Posts composer
The reader-facing homepage strategy depended on an authoring model writers could actually use inside Chorus.

Role

Principal Product Designer

Embedded from Coral on Editorial Tools during the relaunch window.

Collaboration

PM, engineers, editorial leadership

Partnered across CMS and audience product teams during a public relaunch window.

Platform context

Shared Vox Media CMS

Chorus powered major Vox Media properties while composition moved from Ractive to Vue.

Overview

The challenge was a new editorial rhythm inside an old publishing posture

About the product

Quick Posts was a new short-form entry type inside Chorus, Vox Media's CMS for properties including The Verge, Vox.com, Polygon, SB Nation, Eater, and New York Magazine.

The challenge

The Verge wanted more fast, voice-driven posts on its own site. Chorus was built around article production, so even short updates inherited the mindset and weight of a full CMS workflow.

My design response

I advocated for a purpose-built composer instead of a stripped-down article editor, then shaped the MVP around fast authoring with the platform still handling routing, metadata, embeds, and moderation hooks.

The hardest part was not drawing the composer screen. It was getting everyone aligned on what kind of object a Quick Post was.

Where I came in

I stepped into Editorial Tools during a coverage gap

I joined the Editorial Tools work from Coral during a moment when the team needed design coverage. A designer had recently left, and my Coral work had moved into performance improvements, technical configuration, and design system support, which gave me enough availability to help.

That mattered because Chorus was not a side system. It was the CMS editorial teams relied on every day, and the relaunch work depended on new functionality inside it.

Quick Posts was one part of that need. Other editorial tooling work, including experiences around live stories and related publishing workflows, also needed design attention.

My first job was to become useful quickly

Understand The Verge's homepage and editorial ambition.

Understand what Chorus already assumed about content.

Learn where the Editorial Tools PM saw scope, risk, and launch pressure.

Talk with engineers about implementation constraints.

Identify where the ongoing Vue rewrite created room to improve the system instead of forcing a new story type through old patterns.

Alignment problem

Quick Posts touched more than one team's definition of success

Quick Posts did not begin as a simple Editorial Tools request. The Verge editorial team had already been working with the reader-facing product experience design team on what Quick Posts should look and feel like on the redesigned homepage.

Editorial Tools then had to answer a different question: how does a writer actually create one of these inside Chorus?

Alignment map

Team

The Verge editorial team

What mattered

A homepage that felt alive, timely, and voice-driven.

Authoring implication

Quick Posts needed to feel more like posting than filing an article.

Team

Reader-facing product experience design

What mattered

How Quick Posts appeared in the new homepage stream.

Authoring implication

Chorus needed to support the visual and editorial intent of the homepage designs.

Team

Editorial Tools PM

What mattered

Scope, reliability, CMS fit, and launch timing.

Authoring implication

The authoring model needed to stay focused and shippable.

Team

Editorial Tools engineers

What mattered

Implementation quality inside Chorus during the Vue rewrite.

Authoring implication

The design needed reusable primitives, not a fragile custom surface.

Design translation

Connect homepage ambition, CMS rules, PM scope, and Editorial Tools engineering constraints.

Working authoring model

A composer that fit the homepage goal, the CMS rules, and the implementation path.

My role was to connect those concerns and drive alignment around a model that supported the desired output without overcomplicating the authoring experience.

It was not an article. It was not a tweet. It was structured CMS content that needed to feel like posting.

What I found

The CMS was setting the writer's mindset before they typed

The proposed path was to clone the existing article editor and hide the irrelevant parts. That would have saved some build time, but it would not have changed the feeling of the work. A writer opening Quick Posts needed to feel like they could publish a thought, not file a smaller article.

The homepage needed a new supply line

The reader-facing strategy depended on a steady stream of short posts from Verge writers. If the authoring tool felt heavy, that strategy would collapse back into Twitter and Slack.

A smaller article editor was still an article editor

Removing fields from a long-form composer would not change the writing posture. The surface needed to invite a thought, not a miniature story package.

The platform was moving underneath us

Chorus composition was being rewritten from Ractive to Vue while Duet was replacing the front end. The design had to serve the relaunch and fit the future stack.

System model

The editor was only one node in the publishing system

Quick Posts looked simple on the surface, but each entry still had to behave like real Vox Media content. It needed a byline, a URL, a group, a publish state, a home in the Storystream, and enough structure for downstream systems to understand what had been published.

Conceptual Quick Posts system

Writer-facing need

Write a quick thought

Chorus still had to handle

Create a structured entry.

Design implication

Design a composer, not a reduced article editor.

Writer-facing need

Add a reference

Chorus still had to handle

Support images, links, and embeds.

Design implication

Limit the model so the post stayed focused.

Writer-facing need

Publish quickly

Chorus still had to handle

Preserve publish state and routing.

Design implication

Keep the fast path visible and lightweight.

Writer-facing need

Place the post correctly

Chorus still had to handle

Connect to groups, authors, URLs, and homepage surfaces.

Design implication

Use compact controls and sensible defaults.

Writer-facing need

Maintain editorial trust

Chorus still had to handle

Preserve metadata and downstream system requirements.

Design implication

Let the platform do serious work without making the writer manage all of it.

The goal was not to remove complexity from Chorus.

The goal was to keep complexity from becoming the writer's starting point. If too much CMS structure appeared upfront, writers would feel like they were filing an article. If too much structure disappeared, the post would not behave like real Vox Media content.

Decision 01

Design a composer, not a reduced article editor

The core model was two fields. The primary field invited the thought: "What's on your mind?" The secondary field let the writer "say more" with a small set of formatting options. That split created the hierarchy the homepage needed without forcing every post to become a headline plus article body.

Quick Posts composer with author byline, primary text field, secondary text field, image and link attachment actions, group search, and publish button
The focused composer: one screen, two text fields, one attachment area, and a visible publish action.

Product move

A separate short-form composer with language and layout tuned to posting.

Tempting shortcut

Clone the article editor and hide fields that Quick Posts did not need.

Why it mattered

The article editor carried the wrong posture. The goal was not less CMS. It was a different kind of publishing moment.

Decision 02

One attachment kept the post from becoming a package

A Quick Post could include one image, one embed, or one link. That limit was not just scope control. It made the writer choose the object of the post and kept the composer out of full-article territory.

The attachment rule carried product and system meaning:

Image

A visual post could lead with one piece of media instead of becoming an article gallery.

Embed

A tweet, YouTube video, or other supported embed could be the object of commentary.

Link

Internal Verge links and external links could carry different presentation rules without exposing that complexity first.

Early Quick Posts composer sketch showing a short entry form and optional embedded content controls
Early exploration of the posting surface. The final version kept the same core idea: short text, optional context, one attached object.

Product move

Allow one image, embed, or link as the object of the post.

Tempting direction

Support every homepage variation directly in the composer.

Why it mattered

The composer needed enough flexibility to support the homepage vision, but enough constraint to keep Quick Posts from becoming another article workflow.

Decision 03

Let the product need evolve the design system

Quick Posts exposed a weakness in the existing Chorus design system. The available components had been shaped around long-form editorial workflows, but this story type needed to feel lighter, more flexible, and more direct.

The Vue rewrite turned a constraint into an opening

Existing constraint

Rigid article-era components

The system had been shaped around fuller editorial workflows.

Product need

A lighter composer

Quick Posts needed controls that supported short text, optional context, and one attached object.

System opportunity

A reusable Vue primitive

The rewrite created room to improve the authoring model instead of forcing a new story type through old patterns.

This was the design system work inside the product work: use the feature to move Chorus toward authoring patterns that could better support different editorial behaviors.

Product move

Advocate for a flexible authoring primitive during the Vue rewrite.

Too narrow

Treat Quick Posts as a one-off surface with bespoke controls.

Why it mattered

The relaunch needed speed, but Chorus also needed patterns that could survive past one brand launch.

Outcome

The composer helped make the homepage strategy operational

47%

increase in loyal users after The Verge relaunch, cited by The New Yorker

Company-reported metric

15%

overall readership increase from January to September 2023, reported by Adweek

Adweek

Quick Posts was not the whole relaunch. The public-facing change was a more active, voice-driven homepage. The hidden dependency was an authoring model that made those posts practical for writers to create inside Chorus.

The broader relaunch was later associated with stronger reader engagement, including company-reported increases in loyal users and overall readership. I would not frame Quick Posts as the sole cause of those results. I would frame it as enabling infrastructure: the authoring tool that helped the editorial rhythm become possible in daily practice.

Reflection

Publishing tools have a tone

A blank article editor asks for one kind of writing. It says: compose, package, polish, publish.

A posting surface asks for something else. It says: react, share, comment, move.

Quick Posts was about designing that difference into Chorus without pretending the CMS no longer mattered. The tool still had to produce structured content. It still had to respect the platform. But the writer's first experience needed to feel like posting, not paperwork.

What's next

Let's talk

I'm open for new work: whether that's a full-time role, a freelance engagement, or something I haven't thought of yet. If you're working on something where 10 years of designing (and building) for high-stakes environments would be useful, I'd love to hear about it.