← All work

J.P. Morgan Markets Feed

Designing for the workflow, not the homepage.

Client
J.P. Morgan
Sector
Capital Markets
Role
Senior UX designer
Year
2023
Delivery
Homepage launched December 2023

TLDR

J.P. Morgan’s client portal was getting a homepage feed, meant to surface market colour and lead clients to their next tool. I led its UX to a December 2023 launch, and testing found the most frequent clients never came through the homepage. I proposed two other ways in; business kept the homepage for v1.

Independent prototype, built after the engagement. Not J.P. Morgan’s interface.
A trading workspace: FX spot prices across the main column, an execution blotter beneath them, and the market feed docked in the right rail, each feed item carrying save, share and open controls. Open the prototype Loads here, not in a new tab

01

The brief

The homepage feed was meant to bring research forward and lead clients to their next tool.

Clients had a deep library of research and commentary, and a portal that made them browse for it. Some wanted it brought to them; others wanted to reach it fast. The business also wanted the homepage to lead clients to tools they weren’t using. I led the UX with three product managers and a lead UI designer, to a fixed launch date.

The optimised client journey map in three stages, accessing the portal, content engagement and application discovery. Five steps each show the user objective above and a business opportunity rated below; the last step, discovering relevant applications, is rated high for cross-selling products to clients.
Optimised journey: user goal above, business opportunity below Open full size →

02

What shipped

A feed built from what clients had told us, curated beside live, with a way through to the tools.

The feed drew on each client’s research subscriptions and stated interests rather than guessing from behaviour. Editors’ picks sat beside a feed that updated through the day. Recommendations pointed clients to tools, from what they already used and what would make their work easier. I checked the build against the specification before it launched to a selected group.

Launched layout, redacted for the NDA

A greyboxed wireframe of the launched homepage: a curated column of picks with a call to action on the left, a tabbed feed of articles with source and age in the middle, and recommendations with images on the right.
Editors’ picks, the live feed and recommendations, left to right

03

What testing found

The clients worth reaching most were never on the homepage.

Pinning a tool got harder to find, and it shipped that way. The homepage wasn’t where clients expected market news, and many frequent users bypassed it, so the feed and its recommendations sat on a path they didn’t take. I proposed a standalone feed app and a feed-first surface. Business kept the homepage for launch and said it would revisit bypass in phase two. I left before then.

04

Afterwards

I built the direction I’d argued for, to test it rather than describe it.

On desktop the feed docks beside the trade ticket, and a tile opens the full article without leaving the task. On mobile it’s a reading view. I’d test first whether clients find relevant content faster without their main task getting harder.

A feed tile opened as a full article in the prototype: a breadcrumb back to the feed, the headline, a summary, the attributed source, the body in full, tags, and save, share and trade controls in the top right.
Feed tile opening a full article within the prototype
A full story open on a phone in the prototype: back to feed, save and share along the top, the headline, a summary, the attributed source, the article body, and a bottom bar for execute, markets, research, indices and portfolio.
Mobile article view, with save and share above the story
Ask me about it →