We gave NOUR a voice. People gave her their challenges.

There is an easy way to introduce an AI avatar to a consultancy website. Give it the brochure, teach it the navigation and ask it to explain the services. A receptionist with unusually good availability.

That was never the most interesting possibility for us.

When we introduced NOUR to MenschWorks, we wanted to explore a different question: could a conversation help someone discover what they actually wanted to ask?

Our first review suggests there is something worth exploring. It also gives us several reasons to remain curious rather than declare victory.

The interesting part came after “What do you do?”

The most common theme in the review was straightforward: understanding MenschWorks. What do we do? How does it work? What would we deliver?

Those are useful questions. But in some conversations, they became the starting point for something more specific.

Visitors raised difficult feedback conversations, employee engagement, frustration with organisational change and doubts about the role of AI in human development. They also asked practical questions about cost, organisational size and what working with us might involve.

These are themes from the initial review, not a description of what every visitor experienced. Even so, they point towards an important distinction.

Someone asking about leadership development may not need another definition of leadership. They may be trying to work out how to have a conversation they have been avoiding.

Someone asking about organisational intelligence may be less interested in a diagnostic tool than in whether their employees will tell them what they really think.

The useful conversation begins when the category gives way to the challenge.

A good follow-up beats a polished monologue

The review found NOUR most promising in longer exchanges, particularly when she asked a thoughtful follow-up rather than immediately presenting a solution.

In a leadership discussion, the conversation moved towards a specific communication difficulty and the possibility of rehearsing a different response. In exchanges about organisational listening, questions about employee perspectives led to questions about whether people would engage with AI at all.

That second question matters. The right response to uncertainty about AI is not automatically more AI.

The review describes NOUR explaining human-led alternatives, including interviews, listening circles and workshops. This is consistent with how we want MenschWorks to work: begin with the human situation, then consider which approach might help.

An avatar can open a door. It should not insist that every room behind it contains more technology.

An opening, not an outcome

It would be tempting to turn these conversations into a tidy success story. Visitors arrive with problems. NOUR offers insights. Everyone leaves transformed.

The evidence does not support that story.

Our review covered 25 recorded sessions between 9 and 13 October 2026. Twelve contained visitor input; thirteen contained no recorded visitor input. The sessions included short messages and apparent tests or demonstrations. They do not establish 25 unique visitors.

We also do not know whether the conversations led to better workplace decisions, lasting behaviour change or new client relationships. A positive response within a conversation is not the same as a verified outcome afterwards.

What we can describe is narrower, and still interesting: some conversations moved beyond descriptions of services into exploration of a real organisational question.

NOUR did not solve those organisational problems. The conversations sometimes helped make the questions more specific.

That is a starting point for learning, rather than a conclusion about effectiveness.

Listening means taking hesitation seriously

Thirteen sessions contained no visitor input. That number deserves attention, but not a convenient explanation.

It could reflect curiosity, testing, an interruption or difficulty getting started. The review cannot establish the reason for each silent session.

One useful next step is to test clearer invitations. Does someone want to explore what we do, bring a challenge or try a short exercise? Giving people a starting point may be more useful than expecting them to invent the perfect first question.

There is a wider lesson here. Making a conversation available does not mean someone wants to have it. Nor does speaking to an avatar automatically create trust or psychological safety.

People need to understand what the technology is, what happens to their information and how to reach a human instead. Our Privacy Policy explains the recording, retention and uses of NOUR conversations. Please avoid confidential or sensitive information when using her.

Fluency is not a licence to invent

The review also identified something less comfortable: NOUR sometimes described specific programmes and delivery history that the transcripts did not substantiate as established MenschWorks offerings.

That is not a harmless detail. A convincing description can sound like a promise, even when it should have been presented as an illustrative possibility—or not offered at all.

The lesson is clear. An avatar needs reliable boundaries around services, pricing and experience, as well as a personality that makes it engaging to talk to. Where information is unverified, the right response may be to acknowledge uncertainty and introduce a human.

A better handover is another priority emerging from the review. Questions about cost, suitability or delivery are opportunities to offer a relevant next step, rather than continue explaining indefinitely.

These are improvements to pursue, not changes we claim this initial review has already achieved. Information about an actual engagement should always be confirmed directly with MenschWorks.

More room for a human conversation

The most useful question is not whether NOUR can repeat the website. It is whether she can help someone arrive at a more useful conversation with the people behind it.

Our early review gives us reasons to keep exploring that possibility. It also reminds us that good conversational technology needs accuracy, transparency, thoughtful questions and a clear route back to a human.

We gave NOUR a voice. Some visitors brought questions we could not have anticipated in a navigation menu.

The next step is to listen carefully to what that teaches us.

Not a substitute for human conversation. Another possible way into one.

This article draws on an initial conversation review for 9–13 October 2026, rather than an independent evaluation of NOUR’s effectiveness. It shares aggregate counts and general themes, not raw transcripts, visitor identities or testimonials. The small sample should not be treated as representative research.

Dr Marcus Gottschalk Founder, MenschWorks