Data visualisation for the Migration Policy Institute: rebuilding two flagship immigration data tools

The Migration Policy Institute (MPI) is one of the leading centres for immigration research and analysis in the United States, respected in policy debates for its depth and rigour, and well known for its data and data visualisations.

As part of a complete overhaul of MPI’s website and digital offering, they asked us to rebuild their two flagship data tools — the State Immigration Data Profiles and the Unauthorized Immigration Population Profiles — to make them more user-friendly and impactful.

The rebuilt tools are now quick and clear, with a Lighthouse score of 95+ on accessibility and performance, and more user friendly for visitors and for MPI’s data managers.

The importance of clean, manageable datasets

The State Immigration Data Profiles and Unauthorized Immigration Population Profiles had worked well for several years, but the datasets behind them lived in large, complex spreadsheets. Updating the data was cumbersome for the MPI team, and the files weren’t formatted to be read cleanly in code.

The MPI team built new, normalised spreadsheets to our specifications: clean datasets, optimised for parsing into the new Drupal content management system as simple CSV files. For the front end of the tools, we added a ‘wrapper’ within Drupal, using the React JavaScript library to create a frictionless, app-like experience.

This lifted some pain points for the MPI team. A single ‘source of truth’ for each tool means the data can be updated quickly and with far less risk of error, and reused across the tools, charts and embeds without being reworked each time.

Large datasets need clear user journeys

These large, granular data sets needed to serve both expert researchers and general or media users from the same interface. Our approach to building clear, functional tools was informed by regular, detailed conversations with the MPI team, and by the extensive stakeholder research and user testing that we were doing for the main website redesign.

This iterative discovery and feedback process grounded every design and development decision in sound strategy and test feedback.

Previously, the tools relied on simple maps to navigate between US state-level data, which then displayed in large, unwieldy tables. Neither the maps nor the tables were effective on smaller screens, which is increasingly how audiences are working with these tools. Our updated architecture offers much clearer journeys through the data, letting users navigate, digest and interact with it more intuitively.

Moving standard data visualisation to Flourish

Alongside the two bespoke data tools, MPI also needed a better way to produce its more standard data visualisations.

The Tableau data visualisation platform was not performing well for MPI: inaccessible on the front end and slow for the team to produce. As part of the website overhaul, we suggested moving to the Flourish platform to help solve these issues.

We created a suite of 10 branded templates in Flourish, covering the most common graphics types, that could be applied quickly and easily across MPI’s many outputs. They are interactive, attractive, quick to produce and update, embeddable and highly flexible. Most importantly, they are clear and user friendly, adding authority and consistency to MPI’s research and data storytelling.

For more on the thinking behind the templates, read how we approach designing a data visualisation system that still works when the data — and the people using it —change.

Data tools that are fast, accessible and built to last

The new MPI website achieved strong scores across a range of important metrics, and these two tools were a significant factor in that. Accessibility and performance, measured by Google Lighthouse, both reached 100%, and the sustainability score improved markedly, reflecting a rigorous approach to web best practice.

For an organisation whose reach and credibility depend on getting evidence in front of as many people as possible, these scores matter. Strong accessibility means the tools work for users on screen readers or slower connections, and helps MPI meet standards that funders and institutions increasingly expect. Strong performance keeps pages fast and users engaged, and makes these data tools easier to find in search.

A sustainability score reflects how energy- and data-efficient a page is to load. Lighter pages draw less power, use less of a user’s mobile data and load faster on slower connections. It’s a growing consideration for organisations, and it tends to move in step with performance, so the same work that made the tools fast also made them lighter.

Together, these improvements help the site perform in search and give MPI’s users a better, more engaging experience.

If your data tools are straining under the weight of the data behind them, this is exactly the kind of work we love getting into.