Two weeks from brief to production launch
A national publisher's audience was being primed to use this platform on a specific date. Every scope decision had to be made against the calendar, not the wishlist.
Case study - 03 / 06 - Shipped under deadline
A nationwide platform for local businesses, built in 14 days.
When COVID-19 shut Main Street, USA TODAY needed a way to point a national audience at the local businesses still open. We designed and launched Support Local - a discovery platform with map search, business categorization, and gift-card purchase paths - from brief to live in two weeks. It was picked up by The Washington Post and a stack of national outlets the week it launched.
Pandemic response - 2020
supportlocal.usatoday.com ->
PUBLISHING - 2020
CASE STUDY - 14-DAY SPRINT
Brief to launch
on deadline
14 days
Design, build, integrate, ship - start to public production.
Coverage
USA-wide
national
All 50 states. Map-driven discovery for every ZIP.
Press pickups
launch week
WaPo+
Featured by The Washington Post and additional national outlets.
Revenue path for SMBs
native
gift cards
Direct purchase flow for businesses that needed cash, fast.
// 01 - the brief
Turn a crisis brief into a live national platform.
March 2020. Lockdowns rolling state-by-state. Restaurants, shops, gyms, salons - every Main Street business in the country was suddenly cut off from its foot traffic, and there was no national, neutral place a reader could go to find out which of them were still open, still selling gift cards, still hanging on. USA TODAY decided to be that place. The brief came to us on a compressed timeline most agencies would have refused: design, build, and launch a national local-business discovery platform in two weeks.
The platform had to be national in coverage but personal in feel - a reader in Cincinnati needed to find their corner cafe the same way a reader in Tucson found theirs. It had to support map-based discovery, category filtering, and a direct path for buyers to send money to a business via gift cards. And it had to hold up the day it launched, because USA TODAY's editorial coverage was already cued to point at it.
We scoped it to what could actually ship in fourteen days, picked the integrations that gave us national business coverage out of the box: Google Places for location intelligence, an editorial submission lane for businesses to add themselves, and a v2 list for everything outside the core loop. The platform went live on schedule. The Washington Post wrote it up. Other national outlets followed.
// 02 - the challenge
A public launch date with no room for drift.
National coverage on day one
USA TODAY did not have an in-house local business directory. We needed scalable location intelligence and credible business data for all 50 states ready the moment we shipped.
Map-based discovery as the primary UX
The platform's whole pitch was find businesses near you. Map performance, marker density, mobile gestures, and category filtering had to work on launch day at unknown scale.
Google Places API integration under load
Live location lookups, place details, and business categorization came from an external API with quota limits, latency variance, and rate-limit behavior that had to be cached and degraded gracefully.
Gift-card purchase paths for unknown SMBs
The point was to send real money to real businesses, fast. The UX had to connect purchase intent to whatever gift-card surface a given business already had.
Unknown traffic curve from national coverage
USA TODAY editorial was going to drive the launch. The Washington Post and other outlets picked it up the same week. There was no historical baseline to provision against.
// 03 - approach
Every decision had to protect the launch date.
We kept the scope small enough to ship and strong enough to matter.
Two weeks does not leave room for layered approvals, sprint ceremonies, or let's revisit that on Monday. We collapsed the engagement into one war-room channel, scope-froze on day one, and made every subsequent decision against a single question: does this make launch? Anything that did not went into a v2 list.
A
Speed-to-market over unnecessary complexity
If a feature could ship in v1 or v2, it shipped in v2. The platform launched with exactly the surface a pandemic-period user actually needed.
B
API-first architecture
Google Places carried business data. Our app carried discovery, categorization, and the gift-card flow. No local business database had to be seeded before launch.
C
Mobile-first discovery experience
Most users were going to land from a phone, often from a USA TODAY article shared in a text. Every layout decision started at the smallest viewport.
D
Headroom for an unknown traffic curve
We provisioned for a spike we could not predict by caching aggressively, deferring non-critical fetches, and leaning on a CDN-friendly Next.js render path.
PHASE // 01
Day 1
Brief, scope freeze, architecture call
Single kickoff with USA TODAY. Locked the feature list: map discovery, category browse, business detail, gift-card path. Google Places was identified as the data backbone within the first hour.
scope lock
arch sketch
API selection
launch date set
PHASE // 02
Days 2-3
UX flows + mobile-first comps
Wireframed the four core surfaces: home, map search, business detail, gift-card hand-off. High-fidelity comps were produced while the build started on the shell.
4 surfaces
mobile-first
parallel design + build
SSO design
PHASE // 03
Days 3-7
Frontend build - Next.js + React
Next.js application with server-rendered category and location pages, client-side map interactions, and category-based browse routes.
Next.js + React
SSR for SEO
category routes
component lib
PHASE // 04
Days 5-10
Google Places integration + search backbone
Wired Google Places for business lookups, place details, and categorization. Built a caching layer between the app and the API to absorb traffic spikes and keep quota use predictable.
Google Places
cache layer
category mapping
geo search
PHASE // 05
Days 9-12
Gift-card flow + business detail polish
Built the gift-card discovery and hand-off surface: the user-facing path that lets a reader send money to a featured business.
gift-card UX
biz detail
partner hand-off
SSO live
PHASE // 06
Days 13-14
Load test, launch, watch the curve
Final QA, mobile pass, accessibility check, and smoke test against quota. Coordinated launch with USA TODAY editorial and watched the traffic curve as national outlets picked it up.
load test
coordinated launch
press week
0 P1s
// 04 - stack
The exact tools we used.
Frontend
[01]- Next.js framework
- React ui
- JavaScript lang
- HTML5/CSS3 styles
- Responsive layouts mobile
Rendering / SEO
[02]- Server-side rendering indexable
- Category routes browse
- Geo-aware pages local
- Edge-cacheable perf
Data / API
[03]- Google Places API source
- Editorial feed manual
- API response cache quota
- Geo/category index search
Map UX
[04]- Map tiles display
- Marker clustering density
- Mobile gestures touch
- Geo-bounded queries local
Commerce
[05]- Gift-card surface cta
- Merchant handoff links
- Affiliate links tracking
- Business listing CTA support
Delivery / Ops
[06]- Edge cache + CDN spike
- Quota-aware fetch api
- Mobile perf budget ux
- Launch-day monitoring ops
// 05 - what we built
Three things worth showing.
Feature 01 / 03 - for the reader
Interactive local business map, built around your ZIP code.
The map is the front door. A reader opens the site, the map centers on their location, and nearby businesses show up as category-coded markers. From there it is two interactions to a business detail page and a path to send them money.
- Geolocated map centering with category-colored markers and clustering.
- Category and proximity filters update results as the user pans the map.
- Mobile gesture support tuned for phones first, desktop second.
- Server-rendered category and city pages for day-one indexability.
MAP - NATIONAL
Feature 02 / 03 - for the supporter
Business search engine, tuned for near me intent.
Beyond the map, category-and-keyword search let readers land directly on restaurants, bookstores, salons, gyms, or any local business type in their ZIP. Google Places data sat behind a cache layer that flattened quota use during launch traffic.
- Category browse and free-text search across Google Places.
- Location-bounded queries that respect the current map viewport.
- Cache layer between the app and Google Places to protect quota.
- Editorial submissions for businesses USA TODAY heard from directly.
SEARCH - LIVE
Feature 03 / 03 - for the businesses
Gift-card support system, a direct revenue path during lockdown.
Discovery without a way to act on it would have missed the point. Business detail surfaces promoted the clearest available path to buy a gift card or support the merchant while physical storefronts were closed.
- Prominent gift-card CTA on business detail pages when available.
- Clean handoff to the merchant's existing gift-card or ecommerce surface.
- Listing flags for businesses offering gift cards or curbside support.
- Editorial control to spotlight specific businesses.
GIFT CARDS - LAUNCH
// 06 - gallery
What it looks like in the wild.
01 / Homepage - discover businesses near you
02 / Wide coverage
03 / Easy adding new listings
// 07 - takeaways
Three things we'd do again.
// 01
Crisis-response platforms are won on prioritization, not heroics.
Fourteen days is not work harder - it is decide faster. The single most important act of the sprint was the day-one scope freeze: a short list of features that absolutely had to ship, and a longer list that consciously did not.
// 02
Simple, intuitive UX is what holds up under public-use scenarios.
Readers landing on Support Local were stressed, in a hurry, often on a phone, and often on a slow connection. The interface had to be obvious in the first second: map, category, business, gift card.
// 03
Map + search interfaces dramatically improve discovery engagement.
The map gave readers spatial context they could not get from a list; the search gave them a path when they already knew what they were looking for. We carried this pattern into later discovery surfaces.
// the team
A tight sprint team at RTW working directly with USA TODAY product. One shared channel for the full two weeks. No ceremonies, no status decks - just the launch date.
Project lead
Jeff
Project Manager
Niina
Engineering
Valentin
Engineering
Andrei
Front-end
Rebeca
Design / UX
Nemanja
// got a launch date you can't miss?
We do two-week sprints, too.
Not every engagement is a 14-month replatform. Some are 14-day launches with a national audience already pointed at them. If you have a deadline that other teams told you was impossible - we'd like to hear about it.
