Rethinking personalisation for myFT

A new direction for helping readers discover relevant journalism without taking away editorial judgement or user control

AT A GLANCE

Organisation: Financial Times | myFT
Timeline: Discovery, strategy & design (Sep 2025-Jun 2026)
Focus: Product strategy, personalisation, information architecture, mobile & web
cross-platform consistency, discovery & concept development

THE OPPORTUNITY

myFT is the Financial Times' personalisation experience, where readers can follow topics, authors, newsletters and other interests, save articles, and build a more personalised news experience.

Despite being one of the FT's most valuable products for reader engagement, it had evolved incrementally over more than a decade, resulting in fragmented experiences, inconsistent interaction patterns and an increasingly difficult technical foundation.

Rather than redesign individual features, the project set out to establish a clearer vision and stronger foundation for how personalisation should support FT readers in the future. This work ultimately became the starting point for what we called the Thoughtful Assistant.

CHALLENGES

An increasingly complex product

Discovery was an ongoing process. We reviewed previous research and existing product knowledge, but many of myFT's nuances only became apparent as we explored the experience in greater depth and worked with engineers to understand how different parts of the product actually functioned.

Different readers, different expectations

We were particularly interested in less frequent readers who might visit the FT once or twice a week and use it as a secondary news source. For these readers, myFT needed to make relevant journalism easier to discover.

At the same time, we had to preserve the experience of habitual readers who had built strong routines around myFT. These readers often valued editorial curation and explicit control over what they followed, and were less comfortable with algorithmic recommendations taking over that role. The challenge was to make personalisation more useful without making it feel more intrusive or automated.

A mature technical estate

The product had evolved over many years, leaving a complex technical foundation and limitations around what could be changed or introduced. Design decisions therefore needed to work within today's capabilities while still supporting a longer-term product direction.

Designing across disciplines

We worked closely with research, engineering and data science throughout the project. This was particularly important when exploring how existing recommendation capabilities could be improved and how they might eventually support a more thoughtful form of personalisation.

APPROACH 

We started with a design sprint to understand the product, align around the broader strategy and explore possible directions.

• Reviewed previous research and existing product knowledge
• Mapped the different parts of the myFT experience and identified gaps and inconsistencies
• Reviewed competitor and industry approaches to personalisation
  Worked with research to understand the needs of less frequent and habitual readers
• Collaborated with engineering to understand the existing technical architecture and constraints
• Worked with Data Science to explore how existing recommendation capabilities could better support the new direction
• Developed and iterated concepts with the wider team before converging on the Thoughtful Assistant vision

FROM FILING CABINET TO THOUGHTFUL ASSISTANT

Traditionally, myFT functioned more like a filing cabinet: a place where readers manually organised the topics they wanted to follow. We explored how it could become more useful without taking control away from the reader.

TRUSTED ASSISTANT NOW, THOUGHTFUL OVER TIME

The full Thoughtful Assistant vision depended on capabilities that weren't yet available, so the first phase focused on what we could realistically improve with the existing product.

From passive to proactive

• Move beyond static lists and buried functionality
• Surface relevant journalism at the right moment
• Reduce the effort required to stay informed
• Make personalisation feel useful rather than intrusive

What we could do

• Repackage existing myFT capabilities into a more coherent experience
• Improve engagement with followed interests
• Give followed categories greater visibility
• Improve discovery of saved and previously viewed stories
• Create greater parity between app and web
• Test whether the new direction could improve engagement

Preserve reader agency

• Keep explicit following and reader-selected interests at the core
• Complement editorial curation rather than replace it
• Avoid making algorithmic recommendations the default way of discovering journalism

What wasn't yet possible

• More sophisticated recommendation signals
• Fine-tuned email and push recommendations
• Curated briefings and entirely new personalisation features
• Contextual recommendations based on time, behaviour or situation
• Capabilities outside the existing myFT technical footprint

Build towards the future

• Establish a foundation for richer recommendation signals
• Create a clearer distinction between stable interests and more dynamic recommendations
• Design a direction that could evolve alongside future data, technology and AI capabilities

TESTING & LEARNING 

We tested parts of the new experience through A/B experimentation, including a new navigation model that gave followed categories greater visibility.

What we learned

• The navigation experiment did not produce the expected result
• The two experiences being compared weren't necessarily equivalent enough to isolate the effect of the design change
• New navigation patterns may need more time to settle before drawing conclusions from behaviour
• The result prompted useful discussion about how design experiments are framed, measured and interpreted

OUTCOMES

• Established a clearer strategic direction for the future of myFT personalisation
• Delivered the first iteration of the new design direction into the FT app
  Improved the structure and discoverability of existing myFT capabilities
• Created a foundation for more dynamic, recommendation-led experiences
• Developed a clearer model balancing reader control with algorithmic assistance

REFLECTIONS

• Balancing innovation with continuity. Designing for different reader habits reinforced the importance of introducing new experiences without disrupting what existing users already value
• Discovery is an ongoing process. Working closely with engineering revealed product complexities and technical constraints that continued to shape our decisions well beyond the initial discovery phase.
  
• Collaboration needs structure. Working alongside another designer on a complex project taught me the importance of establishing clear responsibilities, shared ownership and effective ways of working early on

About me

I am a designer based in Bulgaria.I love typography, squares, negative space, and almost anything black and white. When I'm not designing, I read, write, drive, watch documentaries, listen to podcasts, and talk to people about random and not so random stuff. 

EMAIL | LINKEDIN |  INSTAGRAM |  TWITTER  

© 2026 | dessutom