Mockup_06

NBC News Dashboards

Setting the strategic pace for future projects by modernizing a core enterprise tool from the ground up.

Client

NBCUniversal
Sept 2022 - June 2023

Contribution

Lead Designer
UX Researcher
Product Strategist

Partners

Product Analyst
Engineers
Newsroom Stakeholders

When the news breaks, nobody has time to hunt through a form.

During live shows, NBC News teams were abandoning the official system for email, because finding their part of a long, confusing form took too long. I redesigned the Producer Dashboard end to end, increasing daily completed tasks by 68% and cutting transcript uploads from 10 minutes to under 30 seconds.

Impact
at-a-glance

68%

more daily completed tasks per user

10m →30s

to upload a transcript with AI

36%

lift in workflow efficiency, with fewer clicks per task

TL;DR Summary

Problem: Producers worked through long, repetitive forms that made it hard to find their part, compare with teammates or know they were on the right form. Under deadline, teams skipped the system and used email, which broke record keeping.

What I did: Led research, strategy and design for the Producer Dashboard end to end, including AI transcript uploads and built-in upload timers.

Outcome: 68% more daily completed tasks, a 36% lift in workflow efficiency, and transcript uploads cut from 10 minutes to under 30 seconds.

My role: Lead Designer, Researcher and Strategist, working with a Product Owner and engineers, Sep 2022 to Jun 2023.

Where this fits

The Producer Dashboard is one of 7 tools in the Newsroom Computer System, the ecosystem I helped architect at NBC News. This case study goes deep on one product I owned end to end.

Situationer_ProdDash

The problem: the official system was too slow for live news

When a request came in during a live show, a form had to be filled out, passed down the line and completed by several people. Every handoff ran through the same long, vertical form full of monotonous fields.

Producers had to:

  • Hunt for their part of the form, with no way to see which fields applied to them
  • Guess whether they were on the right form at all
  • Struggle to compare their work with teammates, which was vital to getting it right
  • Track upload time separately, by hand or in a disconnected legacy tool, even though their pay depended on it

Under pressure, teams skipped the system and used email instead. That kept the show moving but left gaps in the official record, a real problem for record keeping and legal.

Pasted 2026-10-23 at 11.07.00 PM

Research: too many steps, done again and again

I ran research with producers and newsroom staff to find where the time was going. Four findings shaped the redesign:

  1. Too many steps to complete a single task
  2. Repeated data entry: name, title and contact info on every task
  3. A view that went stale: users reset their workspace every 30 minutes to see accurate incoming work
  4. No signal for what was theirs: users read entire forms to find the sections that applied to them

The insight: producers weren't slow. The tool made them do the system's work, so they had less time for the news.

ECM_UserJourney_graphic_03

Key product decisions

Each decision removed work the system had been pushing onto producers.

Iterations_ProdDash_05

1. AI transcript upload

Decision: Connect Otter.ai to the ecosystem so users upload transcripts straight into the Media Asset Management system (MAM).

Why: Uploading and delivering a transcript by hand took up to 10 minutes, often under deadline.

Result: Transcripts upload in under 30 seconds.

transcriptassociation01
timer-resource-02

2. Live Precision Timers

Decision: Build timers into the upload flow that track how long each film, video or document upload to the MAM takes.

Why: Newsroom workers are paid hourly based on accumulated upload time. They tracked it by hand or in a legacy tool disconnected from where they actually uploaded.

Result: Accurate time records captured automatically, in the same place the work happens.

3. Streamlined workflows

Decision: Cut the steps in a task and stop asking for the same details every time by asking the user to pre-fill common information (e.g., name, title, role, etc.) at the start of each working session.

Why: Users re-entered their own information on every task.

Result: Fewer clicks per task and part of the 36% efficiency lift.

requestbuilder02

4. Collaborative workflow panels

Decision: Surface the sections each person owns, and show teammates' progress side by side.

Why: Users read entire forms to find their part, and couldn't easily compare their work with teammates.

Result: People find what's theirs fast, and handoffs during live shows stay in the system.

workflowpanel_01

5. Customizable dashboards

Decision: A dashboard that stays current with regular refreshes and remembers each user's setup.

Why: Users reset their workspace every 30 minutes to see accurate incoming work.

Result: An always-current view of incoming forms, without the resets.

customdashboards01

Challenges & how I overcame them

1. Twenty-five years of muscle memory

The tension: Core stakeholders and end users didn't want to leave the legacy system. They had worked the same way for 25 years, and even an inefficient, disconnected workflow felt safer than a new one.

What I did: Made the case in their terms—less time on forms, more time on the news. Grounding every change in research, from fixing repeated data entry to stopping manual 30-minute view resets, ensured the redesign solved user complaints without disrupting habits. Through clear A/B tests, stakeholders and users immediately saw the new tool's value, allowing us to launch full prototypes for advanced testing and host group demos demonstrating how workflows became faster, clearer, and easier.

The result: Users moved to the modernized tool, and daily completed tasks rose 68%.

2. Overhauling a product people relied on every day

The tension: SVP leadership wanted to overhaul the entire product, while newsroom teams needed it working every day, including during live shows.

What I did: I proposed a four-phased strategy to roll out design updates systematically and minimize the learning curve. The first phase launched a streamlined Request Builder feature for the single user group struggling most with daily tasks. The second phase centralized this feature for all user groups, while the third phase reskinned the entire tool using our modernized atomic design system. The final phase simplified the forms and tasks themselves to eliminate redundant information, ensuring users could maintain productivity without facing a disruptive adjustment period.

The result: All user groups continued their daily work uninterrupted while we achieved our business goals to modernize the tool and simplify user workflows.

Big Picture Impact

Producers completed 68% more tasks a day, and transcript uploads dropped from 10 minutes to under 30 seconds.

68% more daily completed tasks

Tasks completed per day by users working on individual projects, after the full redesign

36% lift in workflow efficiency

Faster to find information and finish work, with fewer clicks per task

10 min → under 30 sec

Time to upload a transcript with the Otter.ai integration vs. by hand

ganp-blp
ganp-blp

Beyond the numbers

"It's a lot easier to manage my tasks and complete the work that I'm assigned to do without having to dig for a needle in the haystack"

  • Accurate pay records. Upload time is tracked automatically where the work happens.
  • Work back on the record. Fewer requests were submitted through outside channels like email, and documentation of all requests were centralized in our product ecosystem.

Reflection

The lesson: people don't resist change. They resist losing the workflow they trust. Showing them their own pain points, fixed, did more than any pitch.

What I'd do differently: Measure email workarounds before launch to prove system adoption with concrete metrics. This ultimately proved difficult to track down because users relied on undocumented, ad-hoc email workflows with no official record-keeping.

What I'd build next: Extend AI capabilities beyond basic transcripts to deeply integrate the workflow. Specifically, I would use AI to pre-fill form fields directly from the story narrative, design integrated plug-ins to connect forms with the broader ecosystem, and completely reinvent how work is displayed and surfaced to users.

 

proddash_endcover02
proddash_endcover01

MLSTUDIOS

With a deep commitment to leveraging innovative technology and human-centered design, I strive to craft digital experiences that are not just inclusive and accessible but also sustainable.