All work

Designing a global crisis response platform that reached 85M+ people

The site wasn’t meant to be beautiful. It was meant to be understood by someone running on no sleep who had just fled their home.

Company
International Rescue Committee · 2022
Role
UX Fellowship, Google.org
Team
Only designer on a fellowship of about twenty
Read
7 min
Illustration of a phone with pages of textHalftone photograph of a phone breaking apart between two handsPixel-art icon of a finger tapping a phone
  • 85M+

    people reached across every crisis site

  • <1 week

    to launch a new crisis site, down from months

  • 20+

    countries, including Afghanistan, Iraq, and Syria

  • UN

    "Connect" award at the 2023 SDG Action Awards, won by the IRC’s Signpost program

The short version

Context

When Russia invaded Ukraine, Google.org sent a $1.5M grant and a fellowship team to the International Rescue Committee (IRC), which runs information services for people displaced by crises.

Problem

Every IRC crisis site took months to build. Its users couldn’t wait: they had just fled their homes, and were exhausted, on old phones and slow connections, often reading in a second or third language.

What I did

As the only designer, I built the site around one principle: design for the exhausted mind. Then I turned it into a template so every future IRC crisis site inherits that principle without a designer.

Impact

The Ukraine site reached 5.5 million people in its first six months. New crisis sites now launch in under a week, are live in more than 20 countries, and have reached more than 85 million people.

Business problem

A new crisis, and a new website every time

When people think of humanitarian aid, they think of food, water and medicine. For someone who just fled, information matters as much: where to sleep tonight, which border is open, what papers to carry.

Russia’s invasion of Ukraine on February 24, 2022 set off Europe’s largest displacement crisis since World War II. The information refugees needed was scattered across Telegram channels, Facebook groups and one-off websites, with no way to tell what was true. And it changed constantly: borders closed, free housing ended, legal volunteers came and went. A wrong answer could mean a night outside, or a ride with the wrong person.

People fleeing their homes in Ukraine

Google.org committed a $1.5 million grant and a team of fellows to the IRC, with a brief to build a site for Ukrainians with a local partner, United for Ukraine, in six months.

Press release of Google.org grant to IRC
See the official press release

What it cost

The IRC runs information services like this in dozens of crises at once, but each new site took months to launch. For someone who left home last week, that’s too late.

The users

We were designing for exhausted users

Our users had just lost their homes. We needed to match their reality, not build a beautiful website.

  • 01

    Old phones, slow connections

    Many were on older devices and 2G or 3G connections, rationing data. Every image cost them time and money.

  • 02

    Three languages, uneven fluency

    People moved between Ukrainian, Russian and English. Many refused to use Russian, even when it was their first language.

  • 03

    No attention to spare

    They were making life-changing decisions while exhausted and on the move. Anything that made them think twice cost them.

Solution

The only designer in the room

A series of decisions in the order I made them

I volunteered to lead UX and found out that meant being the UX team: the only designer on a fellowship of about twenty. After my manager needed me back, I did this alongside my day job for six months.

Team photograph in Warsaw, Poland
Members of the team during a trip to Warsaw, Poland

Decision 01 of 07

Designed for the exhausted mind

A stark direction over a warm one.

Photography is warm and human, but on a slow connection, every image cost our users. So I set one principle for every decision: design for the exhausted mind.

The expected direction

A photo-rich site.

It’s what humanitarian sites look like.

My call

Cut non-essential imagery.

Keep pages light enough to load on a bad connection.

  • Passed over: Higher data usage design

  • Chosen: Extremely performant design

Design explorations with working navigation menu

Decision 02 of 07

Designed a language that didn’t depend on words

An icon that could be understood before the label was read.

Language was the hardest constraint, so every core service, from housing to legal aid to financial support, got a simple icon. I built on familiar conventions rather than a custom set, and in testing, refugees said the icons looked like the ones already on their phones.

Type got the same treatment. I chose Noto Sans, an open-source typeface that rendered Ukrainian, Russian and English cleanly in one family.

The constraint

No single language worked for everyone.

Translation alone couldn’t carry the site.

My call

Icons as the primary way of communicating.

  • Temporary housing
  • Psychological support
  • Legal support
  • Education
  • Childcare
  • Healthcare
  • Immigration documents
  • Temporary housing (alt)
  • Financial support
  • Jobs
  • Social support
  • Pets
  • Volunteer
  • Donate
  • Non-Ukrainian refugee info
  • Other information

Noto Sans

Glyph

Characters

Latin, Cyrillic, Greek and Devanagari letters, all in Noto Sans

Decision 03 of 07

Took the rest of the foundations off the shelf

Familiar patterns meant less for exhausted users to learn.

Fewer custom pieces meant faster pages and a faster engineering team.

Two weeks after I joined, we launched the first version. It was rough and thin on content, but people could use it, and it improved every day.

The constraint

There was no time to build a design system.

My call

Ant, a large open-source system.

Customized only where the experience needed it.

Ant design system

Decision 04 of 07

Created a daily feedback loop with real users

It looked clunkier, and it worked.

Our researcher was in Warsaw, close to the people we were designing for. I’d send her designs at the end of my day, she’d put them in front of refugees, and I’d wake up to findings. The time difference made it a daily loop.

The details it caught were small and decisive.

What testing found

People didn’t recognize the hamburger icon.

When I added the word "Menu," they still didn’t tap it.

My call

Draw a button around it.

On a normal product, that’s a polish ticket. For someone looking for a place to sleep tonight, it’s the difference between finding the housing page and giving up.

Refugees called the site simple and fast; some said it reminded them of Facebook. Where nobody knew what to trust, familiar read as trustworthy.

Research participants
Video feedback from users of the site, blurred for privacy

Decision 05 of 07

Spent the second half on the IRC’s problem

Each built by hand over months.

Our IRC partners kept comparing the Ukraine site to their older sites for Syria and Afghanistan, each built by hand over months for users just as exhausted and short on data. That was the bigger problem.

A week-long design sprint in New York with Google.org and the IRC produced a crisis starter kit: a brand, a social media presence and a website, deployable in under a week. I owned the website.

The brief

A site for Ukrainians, in six months.

Two months in, it was 80 to 90 percent done.

My call

Spend the remaining four months on the IRC’s bigger problem.

Map of areas the IRC Signpost program operates

Decision 06 of 07

Separated what every crisis shares from what’s local

A template that only fits Ukraine isn’t a template.

So I audited the IRC’s most representative sites myself. The surfaces differed by country, but the content came down to a few recurring types.

The fixed core

Urgent information, legal documents, housing and financial aid.

The inputs

Partner, colors, language, content.

Then I poured the existing sites into the new structure. With minor tweaks, they fit, including live services like Refugee.Info and the IRC’s Iraq site.

  • Components of global crisis response platforms
  • Components of global crisis response platforms
  • Components of the Ukraine crisis response website
Applying the system to other response platforms: Afghanistan, Iraq, Ukraine

Decision 07 of 07

Made the principle the default

Nobody has to argue for the principle again.

The Ukraine decisions made this possible. The icons and type carried the most meaning and never needed translating, so they went into the core as-is. Noto Sans, chosen for Cyrillic, also covered Arabic and Farsi.

Before

Every new crisis site took months to launch.

After

A new site doesn’t need a designer.

The IRC pulls colors from the local partner’s logo, loads content into the CMS, and the site comes out light, plain and legible, built for people with nothing to spare.

Results

Impact of the work

Ukraine

The site reached 5.5 million people in its first six months, handled hundreds of thousands of information requests, and helped tens of thousands of people find temporary housing.

United for Ukraine website, mobile and desktop
Mobile components from Ukraine platform

Every crisis after it

The template cut launch time for a new crisis site from months to under a week. It has since been deployed in more than 20 countries, including Afghanistan, Iraq and Syria, and has reached more than 85 million people. The IRC’s Signpost program, which these sites belong to, won the "Connect" award at the 2023 UN SDG Action Awards.

Ben has a rare ability to see beyond the immediate problem and design for lasting, scalable impact. Working with Ben on the Signpost platform was nothing short of a masterclass in thoughtful, user-centered design. Ben went far beyond supporting our collaboration with our Ukrainian partners, he architected a global system that continues to make a real difference in people’s lives, and that truly speaks for itself.

What survives me

The template. Every site deployed since the end of my fellowship has inherited the same core: the icons, the type, and the light pages. None of that needed me to stay.

What I’d do again

Pick one principle and hold every decision to it. "Design for the exhausted mind" settled the arguments about photography, icons, and page weight before they started, and it’s what made the work portable to other crises.

What I wouldn’t

Wait two months to look past Ukraine. The IRC’s older sites were there from the start. Comparing them sooner would have left more of the fellowship for the template.