Coaching is becoming democratised and our tech is enabling it. Could digital coaches eventually replace human ones? Carol Braddick discusses an alternative future

Since the coaching boom of the early 00s, our ‘coaching heroes’ working in HR, talent and L&D have picked their organisation’s spot on the governance spectrum: laissez faire or tight central management? They’ve dealt with pressure to show impact and ROI, and have navigated the maze of coaching certifications and rates.

All their work, though, was done for the select few, such as future global leaders. Thousands of employees didn’t ‘get’ a coach. Instead, they got training, maybe debriefs of their 360 reports, access to an internal coach and whatever remained of their organisations’ investments in manager-coaches and coaching cultures.

These employees are now first in line to benefit from the scaling and customising of coaching, learning and development – at manageable costs. That also includes those working in the gig economy and in self-owned businesses, and clients of life coaching.

A number of players in the UK and US have already jumped in to serve these coaching, learning and development markets. They’ve entered with different services, but share messages of democratisation.

For example:

  • BetterUp: Executive-level coaching for everyone. Ensuring coaching can become not a privilege for a select few but accessible to many
  • Everwise: “Everyone deserves a shot at achieving their full potential. Can you imagine the impact if you did this for all of your people?”
  • Saberr: The world’s first digital team coach, providing always-on coaching support for teams
  • Butterfly: “On a mission to deliver first-rate coaching to all managers in a way that makes sense for the next generation of leaders” (eg, millennials)
  • BoldR: “The AI-powered coach that helps you become better every day”


This new wave of democratisation – of expanded access – has already rumbled through human capital data and HR. The manager who used to wait months to receive engagement survey results for her team now runs a survey via her mobile a few times a year and receives a dashboard of results and action recommendations within 48 hours.

Vodafone recently launched its #digitalfirst, learner-centric Vodafone University, offering “bingeable”, personalised content on demand.

IBMers use an app to give instant feedback, request feedback on themselves and collect feedback on their teams.

This democratisation brings a trifecta of clunky terms: Amazonification, Netflixisation and consumerisation – the latter having already pulled employers into the era of BYOD (bring your own device).

Trifecta is convenient shorthand for the type of experience users should have, and the technology that enables it, namely artificial intelligence (AI).

Is digital coaching – without a human coach – the next step?



Digital coaches could be voice-enabled bots embedded in wearable devices or human-like agents on our screens. They may be a physical object, such as Amazon Echo or Google Home. Clients and digital coaches may hold their sessions using virtual reality or augmented reality using avatars.

Whatever the product, it will get smarter about us through experience with us, because that’s what it’s programmed to do. For example, the digital program we use to check flight arrival times, will become even more skilled at processing natural, messy language from us about our challenges and goals. More advanced programs will also recognise our emotions by processing our facial expressions, tone of voice, pace of speech, vocabulary and, possibly, biometric data.

In organisational settings and uses, it may hoover up and be loaded with some of the information that coaches chase today: stakeholder feedback on clients and analyses of their styles, personalities, diaries and networks.

Its features and capabilities are in the hands of design teams with experts from fields such as: human centred design; human computer interaction; AI; data science; software engineering, design ethics and data protection. They also need subject matter experts (SMEs) in coaching and coach supervision, ie, coaching SMEs.



A coaching SME’s remit should include: the coaching relationship, coaching outcomes and necessary safeguards.

Our coaching SMEs need to help design teams understand that the coaching relationship between two humans, even virtually, is a significant factor in the effectiveness of coaching.

It’s based partly on the empathy, respect and trust that builds between coach and client. Design teams working on chat bots and digital assistants are already experienced at creating interactions that enable users to feel heard and understood.

Our coaching SMEs need to push design teams to create more than this baseline of rapport and trust. In particular, coaching SMEs must help the team understand that effective human-to-human coaching relationships involve support and challenge. They are also the advocates for treating the development of the relationship as part of the client’s growth – on a par to working with the issues the client brings to coaching.

Since digital agents are set up to learn about their users, they may be able to recognise movements in the relationship, such as greater disclosure from the user. Simplifying their complex computations, they assess users’ reactions to their inputs, classify them, eg, positive or negative, and adapt their next move.

Our digital coaches should do more than run these algorithms. In practical terms, this means conversations about what the client finds more and less helpful, the dynamics of coach-client interactions and adapting coaching to benefit the client.

We will run into obvious limitations in what a digital coach can share of its own experience in this new set of interactions. For example, what about that small bit of personal disclosure by a coach that helps build the coaching relationship? Digital coaches might introduce themselves by referring to their coaching experience. But they won’t have any material for small talk.

As a software product with emotion recognition capabilities, a digital coach could discern and acknowledge a client’s emotions. But it won’t have any feelings to share about the experience of working with a client.

Nonetheless, a digital coach could prompt a client to consider how others might be impacted by the client’s emotions around a particular issue. Or a digital coach could function like virtual interviewers that give humans feedback on job interview practice sessions.

The design team also needs to understand the types of outcome the digital coach is expected to facilitate: new behaviours, ways of framing challenges and of relating to people.

In human-to-human coaching, we also aim for the client to gain in self-efficacy and become a better DIY coach. A digital coach can be programmed to monitor these outcomes via self-reports from the client and inferences from coaching conversations. It may also be linked to tools that collect input from others on the client’s progress or other systems, eg, engagement surveys, which may provide evidence of change. We may see hybrid solutions, such as a human coach who uses these tools and data and steps in to assess progress.

Safeguarding issues

What safeguards do we need to build into the digital coach? When my clients get reports from their devices on, for example, their amount of screen time, quality of sleep and degree of multitasking, there’s a common reaction: OMG!

They are shocked by how habitual their device usage has become. They set new goals, not only for less use, but more intentional use. Our coaching SMEs need to speak up for the preservation and growth of the client’s autonomy and agency. The team’s experts in human centred design and design ethics need to show how the digital coach can honour these and be on the alert for unhealthy anthropomorphising of the
digital coach.

Digital coaches must also be designed to be alert to the possibility that some clients need to work with a different specialist, such as a therapist. Algorithms already in use for identification of mental health problems could be included in a digital coach. As we are learning with hiring decisions, algorithms often do a better job than humans. However, the digital coach needs to take a step beyond its algorithms; it needs to be capable of raising this issue effectively.



Is it hard to imagine going from swearing at bots to sharing your aspirations and doubts with them? After all, it’s bound to get us wrong along the way. In June, when Amazon bought Whole Foods, the following joke did the digital rounds: Amazon founder Jeff Bezos had asked Alexa (Amazon’s intelligent digital personal assistant) for 24-hour delivery on organic, free range artisan kale from Whole Foods. Alexa thought he’d asked to buy all of Whole Foods – and did just that.

Despite the likely mess-ups, there are potential advantages of having digital coaches.



You’ll never get an out-of-office message from your digital coach. If it’s anything like IBM Watson Talent, which read my Twitter feed in under a minute and delivered a profile consistent with a traditional, widely used assessment tool, it will be fast.

And its recall is likely both faster and more accurate than ours, especially since most coaches don’t take session notes.


Smart interventions

You find great development recommendations in the last few pages of reports that assessments coaches use. But clients don’t go through their days with report pages 17-21 at the ready. A digital coach will deliver relevant, timely, contextualised tips and reminders. That’s what the more advanced learning and knowledge management tools aim to serve up: right content, right time, right amount.

A digital coach could also be set up for smart non-interventions, eg, to patiently wait through a pause while a client has time to think.


Deeper conversations

Just as we are extraordinarily willing to search the web on topics that we avoid raising with fellow humans, we might also be willing to go far with a digital coach. When digital agents show that they ‘get’ us, our appetite to trust them with more complex tasks and topics may grow. In one study of interactions with virtual humans, participants who believed they were interacting with a computer reported lower fear of self-disclosure and less impression management ( In another, team members working on a task were open to assistance from a bot, taking the practical attitude that if it was there to help their performance, its bot-ness didn’t matter (


Research breakthrough

Will we look back on surveys as being as quaint as floppy disks? Data from digital coaching sessions will be research-friendly: storable, searchable though, unfortunately, hackable. Our main challenges will be data protection and smart use of data volume.

It will be easy, as it is today, to measure activity: picture a report from your digital coaching vendor that crunches the hours of coaching ‘consumed’ by different sets of employees. As we look for evidence of results from coaching, there’s far more data on clients to pull from the continuous feedback, collaboration, knowledge management, productivity, sociometric and social learning tools available today.

Given advances in employee listening such as text mining and continuous listening, it may no longer be useful for an organisation to ask its human coaches for their high-level observations and themes from their coaching sessions. Human coaches are filtering these observations based on the limitations of human recall, client confidentiality and the context of a commercial relationship with their organisational sponsors.

In the future, an organisation might receive quarterly thematic analyses based on metadata from hundreds of hours of coaching. With a flood of data potentially available to organisations, design teams will need to consult with experts in people analytics and business analytics to prioritise how they use these data.


Roles for coaches

There are obvious opportunities for coaches and coach supervisors to support the design teams. The design teams could also use a team coach; think of the challenge of getting to consensus on speed to market. That will require getting the best of the tech sector’s gospel of rapid iteration and the coaching profession’s wish list for evidence-based practices and randomised control trials.

Supervisors could get even busier as their client-coaches make decisions about their coaching value proposition in a more competitive market.

Coaches might pivot their practices to offer hybrid human-digital programs. They can also support organisations as they evaluate and implement new options – again.

For organisational buyers and managers of coaching, it’s déjà vu and plus ça change. People in these roles will still have to do the workforce analyses to make familiar decisions:

  • Which development needs are of highest priority to improve organisational competitiveness?
  • Which talent segments and development needs are best fit for which resource, now that the resources include human coaches, digital coaches, hybrid solutions and Amazonised, Netflixised learning and development, collaboration and knowledge management tools?
  • How do we communicate to the different segments about the new options – bearing in mind the Gartner Hype Cycle of Inflated Expectations and Trough of Disillusionment?


Those in organisations may also be headed for a full rethink of governance. What is a coaching engagement in digital coaching – a monthly use allowance such as with our mobile phone plans? Is governance even possible if coaching is so consumerised that we move from BYOD to BYOC (bring your own coach) and BYOD (bring your own digital assistant)?

Or does governance only apply to whatever use of human coaches continues?

They will also need to work closely with specialists in data protection and privacy to set the T&Cs. Who owns the data from coaching sessions if the company has purchased the software? Or is BYOC the only model that employees would be willing to use? If it is, what T&Cs are product providers asking their users to tick?

As with the data from continuous feedback or internal collaboration tools, there is the question of ownership about and from the employee. What happens when an employee leaves an organisation and wants to take his or her coaching data and employer-provided digital coach out of the door?

We have even more reason to address the supply-demand issues that have already been raised in the coaching market: how many new human coaches can the future market absorb and what expectations should training organisations set about future demand for human coaches?

How will novice human coaches meet the requirements of hours of practice if digital coaching takes off? They could practise with digital clients, but that doesn’t give them any advantage in a market that is potentially even more competitive.

The professional bodies might seek a role in evaluation of digital offers and create demand for a coveted quality seal. However, they may already be behind the product providers.

Digital coaching capabilities are a natural add-on to the more advanced digital platforms for learning and development, collaboration and knowledge management.

The providers of these tools are in prime position to scale coaching to their current user base and leverage their expertise in measurement of user adoption, engagement, satisfaction and outcomes. Providers and buyers may also prioritise other accolades such as awards for best-in-class products at conferences on HR tech, learning and development and the future of work.



The Economist magazine recently tried on names for the demographic that is 65+ but not yet elderly. “Geriactives” was dismissed as uncomfortably associated with senescence, “sunsetters” as patronising. What rang true? “Owls”: older, working less, still earning (

Why the fuss over a name that rings true? Because names matter. They influence our perceptions and, especially when freshly coined, open up new ways of framing emerging developments.

What will our frame be for digital coaching? We can resist it, or shape it for successful democratisation. We can welcome it as a resource for both the many and the few who are facing changes from a related technology: the automation of work.


  • Carol Braddick is an executive coach and management consultant. Carol’s company, Graham Braddick Partnership, designs and delivers coaching and development programs for global leaders and teams. She is a founding fellow of the Institute of Coaching at Harvard University and a mentor on the Coursera Wharton People Analytics MOOC. Her research focuses on uses of technology in coaching such as digital coaching via artificial intelligence, and people analytics tools and apps.  She splits her time between the UK and Arizona.


Try it for yourself

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply