Designing accessible data platforms

Contents

1. Introduction
2. The UX jobs market in 2025
3. My career history
4. Examples
5. The new data paradigm
6. UX roles on Big Data projects
7. Data handovers
8. Designing accessible screens
9. Prototyping tools

Introduction

Data mining and AI are distrupting UX design. This article for UX Glasgow in February 2025 is about a new way of working for designers and researchers.

At the time of writing, I'm sulking with Big Tech. I've avoided using any tools provided by large corporations to create or this article. Take that, oligarchs.

The UX jobs market in 2025

Jared Spool's LinkedIn article Why is the UX Job Market Such a Mess Right Now? — A Comprehensive Explanation has some advice for UX job hunters:

Don’t apply to jobs for which you’re not highly qualified. The “spray-and-pray” approach of applying to every possible job doesn’t work. Neither does applying to jobs you’re overqualified for. These things don’t increase your chances. They just clog up the hiring pipelines for everyone else.

Tailor your résumé, portfolio, and cover letter to the specifics in the job ad. You must demonstrate that you’re 100% qualified for the position. Everything in your application submission must prove that you can do this job because you’ve already done it elsewhere.

If you have a specialisation, it's time to leverage it. In my case, it's designing with data.

My career history

None of the following steps were planned.

• 2006: Data manager for the Scottish Government (temp job)
• 2007: Online Services Analyst for the Student Loans Company
• 2014: Freelance UX Designer
• 2018: User researcher
• 2020: Interaction Designer
• 2024: Pivot: designing with data

When new trends in technology came along, I read about them. You never know when new opportunities will turn up.

Examples of data platforms

Google Analytics offers an interactive dashboard to view how people use online services. Most screens are a mixture of graphs and tables:

The data is automatically updated from a database. When you log in, it will show you a default time range, usually to to yesterday's date.

In the public sector, the Department of Health and Social Care offer the Fingertips website, which allows visitors to find and visualise public health data:

The Observable website shows a lot of fab examples of data visualisations and charts made with the D3.js framework:

The new data paradigm

The government, and all large organisations, hold decades worth of information in various databases.

Designers usually work on user interfaces.

Beneath the UI there's an API layer - the software which allows computer systems to talk to one another.

Underneath the API, there is at least one database. Databases serve up content to users and store their input.

A lot of databases are old. Updating them is complicated and expensive. Modern data-mining techniques make this process a lot easier.

At the same time, users want to be able to access data from multiple sources.

Some 'use cases' include:

• Contact centre staff taking calls can trace a customer and find their record more easily - even if the record is stored on an older system
• Researchers can extract data from audio recordings or images using machine learning, as well as sentiments from text (such as emails)
• Data scientists need to train Large Language Models (LLMs) on good quality data to make them effective

New job opportunities

For years, designers have been digitising paper application forms for their clients and building online services. The UX approach for this kind of project is well-established.

For this reason, the amount clients will pay for these services will gradually decline in future. New entrants to the UX field may have a better chance of winning this kind of work as they often charge less than experienced designers.

At the same time, there are types of design which are becoming more in-demand. For example, the UK lacks affordable housing and adult care staff. The government is investing in technological solutions for these sectors, so expect to find design and research jobs there.

This article is about a specific kind of design which seems to be more in demand than ever: creating data platforms.

UX roles on Big Data projects

User-Centred Design roles have a place in this new area of technology. You may be working with data scientists for the first time.

Data scientists research what sources of data exist in the public sphere; or find out how to buy access to data. They may consider whether the source of data is reliable.

For example, if you're relying on all the schools in Glasgow completing regular surveys on free school meals, some schools might not manage to complete the survey by the deadline. This will lead to gaps in the data.

A data platform project might work like this:

1. A product owner works with business analysts to document what the organisation needs
2. Data scientists discover sources of data and work out what will meet the project's requirements
3. Developers decide how to extract the data, such as extracting it every day using an API
4. Service designers look at the big picture - how the organisation will manage the platform, what staff they will need, how they will work with other organisations to get good data.
5. Content designers find out about how the organisation writes and reviews content, including how to make data easy to understand
6. Interaction or Product designers create a logical flow of screens and design accessible pages
7. User researchers find out what kind of people will use the platform and test out the designs to check they're useful and usable
8. Data Scientists can also create AI features such as summaries and prompts, based on the data you offer users

Data handovers

If you look on LinkedIn, Dribbble or Behance you'll often see polished dashboards with graphs. See the following Data UI example.

If you can produce attractive user interfaces people may hire you to design data platforms.

However, to get to the final product you'll have to consider:

• What data is genuinely important to a user? Why should they use your platform?
• How much control do you give users over the data? Do you only show a summary, or let them build their own queries?
• How do you tell a story with the data, rather than just offering graphs and numbers?
• How do you make a visual medium like a chart accessible to as many people as possible?

You will need to collaborate with Data Scientists to get realistic data to build a prototype. They might hand over a spreadsheet of figures and some suggestions on how to lay out your screens.

Designing accessible screens

The web is built from HTML, CSS and SVG. If you're an interaction designer or product designer, you need to understand these languages.

SVG is the best format for graphics on the web. As an introduction, try the article How to make charts with SVG.

For SVG, you can use the title tag to make the different parts of your SVG image accessible. For a bar chart, you can add a title tag to every bar.

If you hover over the SVG images on this page, a description will appear. This makes each image accessible to people using screen readers.

You have 85% charge in your jet pack:

Prototyping tools for data platforms

Developers are likely to make real data pages using a javascript framework like D3.js. I've found D3 is quite demanding. For my prototypes, I make graphs using SVG.

A quick solution for prototyping graphs might be to use a spreadsheet program, then take screenshots of the graphs.

Or, use an AI solution if you want to rely on a corporation to create your prototypes.

You can adapt the gov.uk design kit to create data dashboards. You can even use nunjucks to create tables with conditional formatting.

Visit my Prototype sandpit for useful code snippets.

You can also see a live demo. Use the password sandpit.

Conclusion

This has been a quick tour of a growing area of UX. This kind of work will suit desigers with an analytical, rational disposition. If you would rather focus on visual beauty this might not be for you. Having said that, designers with an engaging portfolio will find it easier to find work.

At the same time, designers need to consider the ethics of data sharing, privacy and the ethical use of data. Especially at the time of writing (February 2025).

Thanks for reading, any questions?

Back to top or Website home