For years, Learning & Development has wrestled with the same fundamental challenge:

  • How do we help people perform better under pressure?
  • Not just understand a concept.
  • Not just remember a model.
  • But actually execute when it matters.

Whether it’s a sales conversation, a difficult customer discussion, a renewal negotiation, a complaint, or a leadership conversation, the gap between knowing and doing remains one of the biggest limitations of traditional learning.

Recently, I’ve been exploring the concept of the Digital Customer Twin (DCT) and what it might mean for the future of workplace learning.

The idea has been developed in an interesting way by Glen Williamson through his SalesOS model, where Digital Customer Twins are used as behaviourally accurate representations of real buyers and customers. Rather than reading about a customer profile, learners interact directly with the customer before the real conversation happens

The belief here is Do not learn on your customer.

The Problem with Most Learning Interventions

Traditional learning has always relied heavily on knowledge transfer:

  • Courses
  • Workshops
  • E-learning modules
  • Frameworks
  • Role plays

Many of these approaches have value.

However, the moment that really matters usually happens days or weeks later when someone faces a difficult real-world conversation.

That is when habits take over.

A framework remembered in a classroom can disappear remarkably quickly when a customer pushes back, challenges pricing, raises concerns, or simply remains unconvinced.

As Glen writes:

“Knowing what good looks like and being able to do it under pressure are two completely different skills.”

A Digital Customer Twin is designed to go beyond traditional personas.

A persona sits in a PowerPoint deck.

A Digital Customer Twin becomes someone you can have a conversation with.

The learner then has to respond.

Ask the wrong questions and the customer becomes more guarded.

Ask better questions and new information emerges.

The conversation evolves.

In effect, learners practise with a customer before they encounter the real one.

Why This Matters for L&D

When I look at broader learning trends, Digital Customer Twins align with several shifts already happening across the profession:

1. Learning in the Flow of Work

The industry has been moving away from event-based learning for years.

The emphasis is increasingly on:

  • Practice
  • Reinforcement
  • Application
  • Feedback

Digital Customer Twins fit naturally within this model.

2. Deliberate Practice

One of the strongest ideas with the SalesOS is that capability improves through repeated practice in realistic situations.

Pilots train in simulators.

Athletes practise repeatedly.

Yet many customer-facing employees still face their first difficult conversation with no opportunity to rehearse beforehand.

Digital Customer Twins create that rehearsal environment.

3. Personalised Learning

Most AI-driven learning systems personalise content.

Digital Customer Twins personalise interaction.

That is a subtle but significant difference.

Instead of recommending a different course, the system creates a more realistic conversation tailored to the learner’s environment.

Yes I agree it is a different way of learning and practicing and outlines are the potential benefits and some challenges to review.

The Potential Benefits

If implemented well, I can see several significant advantages.

Improved confidence

People enter difficult conversations having already practised similar situations multiple times.

Better behavioural change

Learning moves from theoretical understanding to practical performance.

Safe failure

Learners can make mistakes without damaging customer relationships.

Richer coaching conversations

Managers and coaches gain visibility into performance patterns and development needs.

Scalable practice

Teams gain access to realistic role play without needing multiple facilitators.

The Challenges We Shouldn’t Ignore

As with any emerging capability, there are important questions.

How realistic is realistic enough?

If the customer behaviour is inaccurate, learners may build confidence around approaches that don’t work in the real world.

Data and design quality

The quality of any Digital Customer Twin will depend heavily on the quality of customer insight used to create it.

Human coaching still matters

I don’t see Digital Customer Twins replacing coaches.

I see them amplifying coaching by creating more practice opportunities.

Ethical considerations

As AI plays an increasing role in learning, organisations will need to think carefully about transparency, bias, governance and data privacy.

Happy to discuss and introduce you to Glen, whom I have known and worked with for approaching 20 years!  If this interests you or you want to know, Glen can demonstrate some different DCTs or share your scenario/use case.