Coaching Bots and Automation: How AI Is Reshaping Coaching

You would be forgiven for thinking that creating a coaching bot starts with technology. After all, this is one of the overriding narratives playing out in the industry today.

My inbox is littered with invitations to webinars, talks, sessions and training on using AI in my coaching practice and business. It seems no exaggeration that AI is improving every few weeks, and the different tools and prompting methods are increasing at an exponential rate from automated workflows to dynamic agents initiating tasks on our behalf.

What’s driving this change?

Back in the 90s we had the internet and dot.com revolution leading to websites. Now we have an unfathomable level of investment in AI from the biggest Tech companies.

Coaching is no longer the exclusive domain of executives. It’s possible to coach employees across all levels or roles at low cost and on demand. There are many large-scale coaching providers offering coaching bots to supplement their coaches. The upside of this is there has never been a better time to access coaching services.

For coaches, particularly solopreneur practices, this can feel bewildering and challenging. Trying to stay relevant in a world where some jobs will be replaced and new jobs created by technology is a perceived potential existential threat to the coaching community.

But it’s not all doom and gloom. The latest research suggests human to human coaching is still relevant and, could in fact, be a differentiator for experienced coaches looking to develop their practice.

Purpose and process

After many years working in the IT sector, it’s clear that no technology can create value if the process that sits behind it is flawed. In 50 Top Tools for Coaching we suggest a coaching process to kickstart your thinking about how you might scaffold your coaching method - the lynchpin for what and how you automate.

· Map out your process before you take that first step towards automation

· Identify the key touch points with your client

· Workflow the systems interacting with your coaching method

· Identify the different resources you use as part of your overall coaching approach

When you’ve mapped it out, be honest with yourself about what works well and what could be improved. Identify which parts of your process sit well with technology such as surveys, assessments, follow up emails and calendar bookings, and which parts work better with personal contact.

Clarifying your approach is a key step in designing your AI because the clearer you are about this, the easier it will be to:

· workflow automations such as contact emails

· set up calendar appointments

· identify which applications fit with each other

Experimenting with AI and automation over the past two years has taught me (and is still teaching me): patience and clarity of thinking is more, or, as important than the application you use.

Coaching philosophy and values

If purpose and process are the building blocks, your coaching philosophy and values are the lifeblood. Ask yourself deep reflexive questions about what you fundamentally believe to be true about coaching. Without understanding your own guiding star on what coaching means to you it’s possible you will use technology that is at odds with how you want to present yourself as a coach.

· Privacy

o How will data be stored by you and the AI provider e.g. ChatGPT, Gemini, Zapier

o Will the data be used to train the Large Language Model (LLM) – many paid LLMs make it possible to turn this setting off e.g. ChatGPT, Claude, Gemini etc.

o What privacy standards apply in your part of the world, e.g. GDPR, how will these be met in your use of AI

· Ethics – early experiments with AI found humans create bonds with their chat bot known as the ELIZA effect. 

o Be specific in your instructions - AI is not a human

o Be explicit about how your clients will interact with the AI

Coaching Bots as a Differentiator

There is a commercial reality: organizations and individuals will still crave personalized services.

· Could coaching bots be a differentiator when paired with automated coaching tools?

· How might you leverage AI and access to your coaching?

Coaching bots perform as well, and arguably better, for goal-related coaching assignments. Human-to-human coaching is better for transformation. Humans can read the non-verbal cues - the unspoken dialogue that passes between the coach and client which cannot be replicated by technology. We also use self-awareness as a resource in the coaching process. These are a few of the differentiators we currently have because coaching bots can’t replace our human qualities.

Taking a strategic view of your coaching practice includes developing a service where AI interacts with your business processes to make it easier and scalable.

For those coaches operating as solopreneurs this helps you create a 24/7 operation and focuses your energy on higher value services both for your clients and your coaching practice. In-house coaches can focus their energy on the high value transformational development.

Using technology to deliver coaching sessions and support the coaching process

Last year I embarked on an experiment with my own coaching bot – Sol. My quest was to explore if AI can ethically qualify a potential coaching client. The experiment provided valuable research data for tools and approaches. It forced me to think deeply about how I coach and what ‘magic’ means in a coaching session.

Here are some of the questions from my quest:

Will my technical skills be up to the job?

Whilst it’s true that some coaching bots require little or no technical skills e.g. CoachVox, I decided I wanted to learn more about the ‘backend’ of a coaching bot and chose coachbot.ai.

The key is understanding the platform you’re using and checking it is compliant with privacy and data rules. There are also platforms specifically designed for the health and wellness sector that are compliant with industry standards and regulations e.g. medical.chat-data.com.

How do I create a bot that both demonstrates rapport yet doesn’t pretend to be human?

I used the word experiment to describe my journey, and this is definitely an area where testing comes into its own.

An example is what happens when a client says, “I don’t know?” We can use this to “tell the coach bot” how to respond:

· DO NOT use the phrase: “What if you did know?”

· DO NOT offer a list of suggestions immediately.

· DO normalize ‘not knowing’ (a transitional space)

· DO reframe by asking for one possibility or one small fragment of an idea.

How do I replicate intangible human interactions, like ‘holding space’ in code?

How human you make your coaching bot is an entirely personal choice. I err on the side of caution, given what we know about our human tendency to attribute human qualities to inanimate objects. The ELIZA effect mentioned earlier came out of an experiment with a mock Rogerian psychotherapist created by Joseph Weizenbaum in the 1960s. He found that even though it was a computer program, some people interacted with it as if it were human.

This presents both ethical and design challenges. The aim is to create sufficient rapport and natural conversational flow without the coachee feeling they are ‘talking’ with a human.

Will it really help me qualify serious clients or just be a time waster?

Many coaches are now using coaching bots (e.g. CoachVox), to help qualify serious coaching clients and my experiment is just on the edge of launching.

Building Sol forced me to think hard about my coaching process and be specific about what I mean by concepts like silence in a coaching session, or how I respond to “I don’t know” and to write an instruction for it.

Did I want it to replace me or make life easier for me?

This question goes to the heart of choice: AI replacing coaches versus partnering with coaches. There are things coaching bots do much better than humans. For example, they never get distracted, tired, lead the client or any of the other human foibles that can sometimes get in the way of great coaching sessions.

A coaching bot partnering with your coaching practice means you have a 24/7 coaching assistant to help support clients at the point they need it.

AI can help with:

· between-session accountability and goal tracking

· preparing clients for coaching sessions

You can also create a digital twin with all your knowledge accessible to coaching clients when they need it. It all starts with your purpose and process for using AI.

The next step in coaching bots and automation

Whilst creating a coaching bot doesn’t start with technology, it means we can partner with it to deliver coaching both at scale and highly targeted. There will always be a place for human-to-human coaching and technology can enhance this and help make a difference to your coaching clients.

Realistically coaches who don’t use any form of technology in their coaching practice may get left behind. More organizations and private individuals expect some form of outside hours or ease of access to support.

You don’t need to be a technical genius to create a coaching bot, but you do need to ask some deep questions about what, how and if you’ll integrate technology into your coaching practice.

Save 30% on 50 Top Tools for Coaching with code AHR30.