NBC News · Lead Designer
NBC News App
Balancing new features with a new framework.
Overview
The NBC News app was built on a dated stack with two page templates, few entry points to video, and no tools for editors. Nearly all of its content was auto-fed. Across two releases, I led design for a video-first rebuild, the app's first mobile CMS, and an editorial toolkit that scaled across NBC News, TODAY, and Telemundo Noticias.
Readers, editors, and engineering each needed something different, but we couldn't deliver it all at once. To keep the deliverable list aggressive yet realistic, I partnered with engineering, product, and editorial to identify the features that would most elevate the product while meeting business goals within the proposed timeframe.
Results: Video starts up 51%. Daily active users up 20%. Editors used 87% of the new curation tools every day.
The tension
Readers wanted immediate breaking news, then local news and weather, and a highly performant search. Saved content and history were lower priorities.
Editors were boxed in. They had no curations, automations that repeated the same stories during heavy news cycles, and almost no way to package coverage. During big stories, one second-tier story could spread across several sections and crowd everything else out. The team nicknamed this "too much Will Smith," after the slap story appeared on almost every page.
Engineering had a codebase that couldn't support user accounts without significant refactoring, and a company-wide NBCU account system that wasn't ready to scale into News.
The business needed more video views with new ad inventory to keep the product funded.
The team had 3 product managers, a design manager, 2 designers, 2 researchers, 3 editors, and 2 development managers.
Research
I designed for two users: the reader, and the professional who serves them.
Readers
We tested with NBC News and TODAY app users. 78% of News app users responded positively to increased video access. 62% were very satisfied that they could find everything about a storyline in one place. Market research from our Insights team confirmed that breaking news was still the feature that mattered most. That changed our focus to notification timing, frequency, and an in-app inbox.
Editors
Rather than ask editors what they wanted, we asked them to show us their pain points. They documented every place the tools limited their storytelling, using screenshots, notes, and recordings, and produced 20 pages of pain points. I followed up with interviews, added a competitive audit of where rival apps did better or worse, and sorted every item by storytelling impact, development effort, bugs, and nice-to-haves.
Notifications
Users were already experiencing notification burnout. I mapped the use cases for each alert type, user tested to find where users started to feel annoyed and if there were settings that would make push notifications more appealing. The resulting approach combined geo-targeting, delayed sends for lower-priority alerts, and user-controlled preferences, so alerts showed up when they mattered without becoming constant.
The call
Feature negotiations
The research gave three clear priorities: local news and weather, highly performant search, and editorial storytelling tools. Accounts were the third reader request, and the most expensive one. Engineering reached the same conclusion from the other direction: accounts required refactoring and a company-wide account system that wasn't ready yet.
Pausing freed the team to ship what readers and editors needed most.Account shipped in v2 using the NBCU system, and will support current and future paywalls.
I built the upgrades on tools editors already knew.
The fastest wins needed little or no new API data, so no borrowing engineers from another team, and they used content packages editors already understood. I added new options to five existing components: heroes, promos, topics, sections, and banners. Editors got flagging, mixed content types, and hierarchy variations without having to learn a new system.
I kept stakeholders aligned with a light, regular cadence.
Editorial, product, and engineering met for one hour to set priorities and 30 minutes to agree on solutions, then checked in weekly for 15–45 minutes on progress and roadblocks.
What shipped
Video first
Video got prominent placement on the home screen alongside articles, live streams, and audio. Dedicated sections covered shows and live coverage, with short, medium, and long-form options. An audio option covered broadcast shows for when watching wasn't practical.
The first mobile CMS
Editors could change layouts, rewrite headlines and copy while keeping the originals, reorder and pin stories, and choose whether each component was automated or hand-curated. Automated sections stayed fresh during slow periods, and manual curation took over when a big story broke.
Hero package variations: image size and repeatable elements set hierarchy.
A system built for more than one brand
Every component was brand-agnostic and aligned with Bento, the NBC News design system. The same system updated TODAY and Telemundo Noticias, and we prototyped how Sky News could use it.
Revenue
Ads were built into content feeds, plus pre-, mid-, and post-roll placements in a new video player.
The results
- +51%
- Video starts across the mobile app portfolio (9M → 13.6M).
- +20%
- NBC News app daily active users (920K → 1.1M).
- +25%
- Videos viewed per video visit.
Reflection
What users say they want and what they use are different questions. Nearly 9 in 10 users asked for personalization. When we shipped active controls, fewer than 20% used them. I recommended passive personalization based on history and activity instead, and that became the approach for later releases. Next time, I'd test demand with a lightweight prototype before building active controls.
Check old requests against the new framework early. Editors used roughly 87% of the tools we created. The remaining 13% of curation tools were either built for specific events like breaking news, or were requested features under the old system that the new framework made unnecessary. Reviewing the request list against the new framework before building would have saved that effort.
Connect
Let's talk about what we can build together.
- Design, strategy and component architecture at scale
- AI-assisted builds, from sketch to working product
- Ask me about WCAG, I treat accessibility as a requirement
About
Hello, I'm Lisa.
I've been a UX director, strategist, and AI-native builder for enterprise-level companies with 1.5M daily users, for mid-size trades and nonprofits, and small, solo-founder companies across the West. At any scale I distill input into something that solves for the user and moves the business forward.
I live in the Sonoran Desert, which inspires me daily with its surface simplicity and underlying complexity. I'm grounded and nerdy: ask me about WCAG compliance or anything equestrian and you'll get the same enthusiasm. I'm most proud of raising a son full of gratitude and spoiling some very good horses.
