9 months in. Our strongest week yet.

Welcome to the first issue of The Voice Behind VoiBo.

This is not marketing copy.

This is a practical account of what happens when voice AI meets real recruitment workflows. What's genuinely working. What's failing. What's improving, at the coal face of automated, agentic AI within talent attraction. And what we are learning commercially along the way.

Written by me, Simon, founder of VoiBo. I'll cover a wider selection of automation and AI use cases, with a bias towards voice AI.

If you run a recruitment business, lead talent acquisition, manage recruitment technology, or invest in this space, my aim is simple.

Share the reality, not the hype.

9 months in. Our strongest week yet.

Last week I was away skiing with the family.

Time away creates useful distance. You stop reacting and start evaluating. That's a valuable pattern for me, since the very decision to build VoiBo came to me on a flight to a friend's wedding.

You ask better questions.

• What is genuinely improving?
• Where are the risks?
• What still feels fragile?
• What needs building next?

What was particularly satisfying was coming back to our strongest operational week to date.

Not in theory. In usage.

The data is moving

Across our live deployments, we have seen:

• A 15% reduction in premature call terminations since December
• Early drop-offs now below 5% for some clients
• Our highest ever week in total call minutes processed
• Our highest ever week in total calls completed

That matters.

In voice AI, early drop-off is the real trust metric.

If candidates hang up quickly, it is usually one of three things:

1. The voice feels unnatural
2. The flow feels robotic
3. They do not understand why they are being called, or the call simply isn't relevant to their interests

Over the last nine months, we've invested a lot of time refining those variables.

The improving data suggests two things. The system is getting better through iteration. And the population is becoming more comfortable with voice AI.

We're confident it's both. Either way, the direction of travel is encouraging.

The challenge nobody talks about

Have you ever spoken with an AI voice and had it misinterpret a one-word answer?

Did it frustrate you that it couldn't handle something that simple?

Here is what nobody explains:

One-word answers are often the hardest thing for an AI system to interpret, especially given that all phone lines compress and delay audio.

When candidates realise they are speaking to AI, many instinctively simplify. They assume shorter equals easier. So they default to one or two word responses.

In reality, the opposite is true.

A full sentence provides context and structure. Advanced tools use probability signals to interpret meaning, much like we humans subconsciously do.

A single word removes that context. Ambiguity increases. If the system is unsure, it will usually ask again.

Even within one language, this matters. In multilingual workflows it becomes even more nuanced. Nein or nine? We or oui? This is something we have been refining recently with a forward-thinking multilingual recruitment client.

These operational edge cases shape user experience far more than people realise. It is the sleeves-up work that happens at VoiBo every day.

As society adapts to interacting with different types of AI, this is one of those small behavioural shifts we will gradually get used to.

Worth keeping in mind when you next speak with a digitised voice.

Friends of VoiBo

Many of you have already shown huge support, excitement and belief in VoiBo from the start, as well as offering some great advice. Thank you for that. It makes a massive difference.

To those who are newer to VoiBo, or to voice AI within recruitment. If you are:

• Running a recruitment business
• Leading internal talent acquisition
• Exploring automation in high-volume hiring
• Investing in recruitment technology
• Rolling out adjacent tools

This newsletter is for you.

I will share the wins, the friction points, the data shifts, and the commercial lessons as we build. It will give you a window into the art of the possible, with real-world examples, and the future of recruitment processes.

A favour to ask.

If someone in your network is thinking about voice AI in their recruitment workflow, or is interested in backing responsible recruitment technology, I would genuinely value an introduction.

Recruitment remains a relationship-driven industry.

AI does not change that.

It changes how we use our time within it.

I am hugely enjoying the challenge of helping recruiters become more agile, more efficient, and more focused on high-value conversations with the right talent.

Have a great day.

Simon