Data-Driven Learning Design: Insights from GSK’s Global Head of L&DEpisode Summary

In this episode of Product Design for Learning, host Greg Arthur is joined by Carlo José, Global Head of Leadership and Learning at GSK. Together, they unpack how data, performance consulting, and design thinking intersect to build impactful learning strategies at scale. From debunking mobile learning myths to discussing the power of personas and practical ways to measure impact, Carlo shares rich insights from his global experience at GSK and HSBC. Whether you're a solo practitioner or part of a large L&D team, this episode is packed with actionable advice for designing smarter, data-informed learning products.

Guest Profile

Carlo José is the Global Head of Leadership and Learning at GSK, based in London. With over a decade in L&D and nearly two decades across operations and service delivery, Carlo brings a unique blend of data fluency, leadership development, and user-focused design thinking. Previously, he held senior roles at HSBC and has led high-impact leadership programmes across global enterprises. Carlo is passionate about making data practical, using it not just to inform decisions but to drive meaningful learning outcomes.

Key Take-Aways

• Data is a thread, not just a phase – It should be embedded across the entire learning lifecycle, from scoping to impact measurement.
• 
  
• Performance consulting starts with empathy – Ask better questions before jumping into data or solutions.
• 
  
• Personas are a tool, not gospel – Use them for direction, not as hard rules, and always supplement with live insights.
• 
  
• You don’t need to be a data scientist – But you should be data-literate enough to avoid being misled or overwhelmed.
• 
  
• Leadership development needs time and real-world data – Sometimes the ROI is years down the line, and that's okay.
• 
  
• Polls, pulse checks, and social listening can be just as powerful as formal analytics.
• 
  
• Focus more on outcomes than vanity metrics like attendance or completion rates.
• 
  

Questions Asked During the EpisodeHow do you approach data in the learning design process?

Carlo sees data as both a standalone phase and a continual thread throughout the learning lifecycle. From identifying problems to measuring impact, data informs everything.

How has your background in operations shaped your L&D practice?

Coming from an operations background, Carlo brought a performance mindset to L&D, focusing on measurable outcomes and process improvement.

Is data still as hot a topic in L&D as it used to be?

While AI now dominates conversations, data is still foundational. Carlo explains how sentiment analysis and structured/unstructured data are becoming easier to leverage.

How do you balance qualitative vs quantitative data?

Carlo advises categorising data into textual, numerical, and opinion-based insights, all of which have value at different stages of product development.

How do you stop teams from being overwhelmed by data?

Keep it scoped, work with data partners where available, and use user-friendly tools like ChatGPT and poll surveys for fast insights.

Are personas still relevant in modern L&D design?

Yes, but they’re just one tool. Carlo uses them when available and combines them with user feedback loops like experience networks and shop-floor visits (Gemba walks).

How do you measure long-term impact and behaviour change?

It’s not always instant. Carlo shares an example from HSBC where a CEO pipeline programme showed ROI years after delivery via LinkedIn alumni tracking.

What are your thoughts on Level 1–4 evaluation models?

Use Level 1 and 2 initially, then shift focus toward Level 3 (behaviour change) and 4 (business impact) for sustainable value.

How can small teams do better with data?

Start with free tools like ChatGPT, tap into WhatsApp or Slack communities, and run small polls for direction. It’s about curiosity, not perfection.

Any practical advice for teams starting their data journey?

Use canvases (like Roman Pichler’s product canvas), document what you don’t know, and follow the clues your data reveals. Don't let the perfect be the enemy of the good.

Chapters and Time Stamps[00:00] – Welcome & Introduction

Greg welcomes Carlo and frames the episode around data in L&D.

[02:35] – Carlo’s Career Journey

From HSBC to GSK: how Carlo blends ops and L&D.

[05:01] – Why Data Deserves a Bigger Role in Learning

Why it’s more than just a one-off step in the process.

[07:53] – Storytelling, Surveys & Sentiment Analysis

Data now includes unstructured insights and AI-powered processing.

[11:41] – Performance Consulting & Asking the Right Questions

Where to start when someone says “we have a problem.”

[14:29] – Using Personas & Real-World Testing

Why personas matter and when they don’t.

[21:40] – Dealing with Data Overload

How to avoid drowning in spreadsheets and conflicting opinions.

[29:18] – Are Personas Divisive in Design Thinking?

Why Carlo takes a balanced view on their role.

[34:53] – Quick Data Tactics That Work

Polls, WhatsApp groups, and ChatGPT as everyday tools.

[39:11] – How Far Back Should You Track Data?

The ROI timeline for leadership and long-form learning.

[46:09] – Measuring Behaviour Change & Learning Impact

How programs can evolve from theory to real results.

[53:41] – Justifying Learning’s Existence

If your team vanished, would anyone notice?

[56:01] – Final Tips & Free Tools

Carlo’s practical advice for teams getting started with data.

About the Podcast

Product Design for Learning is where strategy meets storytelling in the world of learning design. Hosted by Greg Arthur, the podcast dives into real conversations with industry leaders who are shaping how learning happens in the modern workplace. Each episode blends product thinking, user experience, and learning innovation to help you design smarter and scale faster.