Posted: 19 Feb. 2020 5 min. read

Automating HR service delivery: Demystifying the conversational AI market

Posted by John BrownridgeGreg VertGautam Shah, Simon Schoon, and Kathi Enderes on February 19, 2020.

Conversational AI is a hot topic for organizations evaluating the next generation of tools to optimize HR service delivery or seeking to augment the work of HR professionals and people managers. Yet many of our clients are taking a “wait and see” approach to a confusing and highly fragmented market. For those willing to learn how to navigate this landscape, there may be substantial opportunities to improve work outcomes for HR and the overall employee experience.

The adoption paradox: Conversational AI in HR

With a 40 percent increase1 in smart speaker ownership in the United States over the last year alone, the pace of adoption of tools that automate conversational interaction is staggering. Many of us have become accustomed in our personal lives to chatting with an assistant to send money, checking our credit score, buying insurance, or making a reservation—sometimes without even realizing we are doing so. More than 30 percent2 of CIOs have deployed conversational AI as a channel to engage front-office customers, and adoption in the back office (notably within the IT function for high-volume transactions like password resets) is increasingly widespread. However, this type of consumer-grade experience is still rarely provided to employees as a means of quickly answering an HR inquiry or taking an HR action.

What explains this paradox? Undoubtedly, HR leaders are faced with a confusing, crowded, and maturing buyers’ market. Virtually every major cloud and ERP provider is investing in tools to automate chat interactions across their product suites, with many making bold claims about what their solution can do. Yet the path to viable conversational AI can be long and winding. It involves an iterative, “fail fast” mindset and leaders moving past common misperceptions and understanding the real-world performance and application of conversational AI. The HR function was an early adopter of cloud and has some of the most innovative technology within the corporate arena; it’s now time to lead with this new opportunity to free up capacity and refocus the work of HR professionals to efforts that drive team and corporate performance whilst reinventing the employee experience.

Let’s start by defining conversational AI

Conversational AI is any solution that combines natural language understanding with analytics or machine learning and automation to orchestrate a business outcome. It differs from other applications of AI; it offers a means to make customers’ lives easier and faster by speaking in their own language, via any channel, to interact with tools, insights, data, and applications.

Yet not all solutions in this marketplace are born equal. In fact, there’s a broad spectrum of capability within the vendor landscape. Supporting complex conversations requires a higher degree of intelligence, and today’s leading solutions are not simply chatbots. Instead, these should be considered digital assistants.

Conversational AI is also not a self-learning technology. Solutions must be “trained,” much like human employees. In contrast to robotic process automation (RPA), it’s more dynamic in its ability to follow business processes and engage directly with customers.

The conversational AI market

There are more than 1,000 vendors3 currently operating in this space. A majority of these offer a rudimentary question-and-answer capability (that is all too often obscured by layers of the latest AI buzzwords within their external marketing). Certain vendors have begun to emerge as leaders, with a proven ability to deliver enterprise-ready solutions across market segments. Yet the complexity of the market continues to be staggering, and understanding what you get for your money has never been harder. Do I need more language control or less? Does the complexity of my use case warrant a more expensive tool? What level of capability do I need from my in-house talent to get value, and should I prioritize speed to deployment or developing a highly bespoke solution that meets the customer experience standards of my business?

Faced with these choices, HR leaders need a framework to understand the current landscape. We see four different types of toolset emerging and break them down below to help support you in your market evaluation:

SOLUTION TYPE 1: Enterprise technology providers

Chat features that are embedded into existing HR cloud, ERP, or CRM technology offerings.

Why would I use this type of tool?

“I want to make the most of my existing HR technology investments, keeping my environment as simple as possible, without introducing extra cost or worrying about automating significant amounts of work.”

What are some potential drawbacks?

  • Some cloud and ERP providers are playing catch-up in capability with product innovation, often lagging behind pure-play providers.
  • The assistant also typically has a more limited ability to integrate and take actions in third-party apps, based on a design that is focused on improving access to services within the native application (with some notable exceptions).

SOLUTION TYPE 2: Platforms

Pure-play conversational AI that offers a full spectrum of capabilities out of the box.

Why would I use this type of tool?

“I want a fully-featured conversational AI solution that can automate a broad spectrum of work across the HR function and other areas of my business. Product innovation is key, and it has to be simple to deploy, integrate, and secure.”

What are some potential drawbacks?

  • Platforms are often the costliest providers, and the business may have limited or no existing relationship, given the assistant is the primary offering of the vendor.

SOLUTION TYPE 3: Frameworks

A “build-your-own-bot” toolset with limited capabilities or features provided out of the box, but the ability to add more services at different license costs and develop simple use cases fast.

Why would I use this type of tool?

“My priority is speed and flexibility. I want to develop proofs of concept at low cost to prove value, in order to ‘fail fast’ and learn.”

What are some potential drawbacks?

  • Applying framework tools to complex enterprise process landscapes often requires strong in-house technical development capability.

SOLUTION TYPE 4: Points solutions

Assistants that have been developed to automate a specific HR domain like talent acquisition, often including predefined language taxonomies and assets designed to accelerate implementation timelines.

Why would I use this type of tool?

“I’m looking to turn the customer experience and efficiency needle in a particular area of my service and want to do this with as little work as possible in-house to be able to achieve this outcome.”

What are some potential drawbacks?

  • There is often limited or no capability to scale point solutions to other HR domains.
  •  The vendors are typically smaller in nature, with few large enterprise customers and less proven integration and security models.

A call to action

Robotic and intelligent automation technologies present major disruption and substantial opportunity for the HR function, and conversational AI is just one facet of a much broader HR automation landscape. Despite the “wait and see” approach being taken by many of our clients, the technology is here to stay and should be recognized as one of the key HR technology priorities of the next one to three years.

At Deloitte, we’ve developed several assets to help you to thrive in your journey to experiment and adopt technologies, ranging from tools to support opportunity assessment to accelerators for different implementation phases. Reach out to learn more.


John BrownridgeGreg VertGautam Shah, and Simon Schoon


1Bret Kinsella, “US Smart Speaker Ownership Rises 40% in 2018 to 66.4 Million and Amazon Echo Maintains Market Share Lead Says New Report From Voicebot,”, March 7, 2019,

Join the conversation