Skip to main content

Why you need to interview chatbot candidates

(Image credit: Image Credit: Alex Knight / Unsplash)

Chatbots are getting a lot of attention for several reasons. Customer service expectations are at an all-time high.  Delivering an exceptional experience and reducing service friction in the support process can be a true differentiator.  Think of it as enabling a digital transformation of your workforce. Today, support teams have an opportunity to upgrade their service desks using intelligent chatbots. These intelligent chatbots can drive real business impact by optimising the blend of machine-driven support and human-based service activities.

It may sound easy but it is not.  There are many different types of chatbots.  Just as you interview people to determine the right level of experience and skills for a role, you need to take the time to analyse and question the different chatbot vendors to determine the best fit for your organisation.

Intelligent chatbots, also known as virtual agents, provide end users with better service experiences and enable service teams to focus on the most impactful activities. Basic or simple chatbots are limited to responding to questions with pre-loaded answers and simple match capabilities.  In order to improve your service management, you require an intelligent chatbot.  These virtual agents can engage in more complex conversations and decisioning about the service required.  The virtual agent should also learn from these interactions to fine tune problem statements matched to intent and skills; present the correct knowledge to a problem resolution intent for support of the end user; or factor in infrastructure (or other events) for dynamic decisioning in the virtual agent skills. 

In some cases, intelligent chatbots can even achieve “I wish I knew” pro-activeness of knowledge on infrastructure, security and other events for end-users.  While this point may seem trivial, the impact can be quite large.  In a recent study by Gartner, the proactive warning approach can knock off hundreds of thousands of dollars in cost avoidance for a mid-size business.

Virtual agents will significantly alter the way employees interact with support desks. There is currently no standard method for evaluating virtual agents.  In fact, several people have asked me how to compare one virtual agent to another.

My advice is to act as you would when hiring a new employee.  Think of a virtual agent as a “digital” employee who needs certain skills and abilities to effectively perform a job. 

Here are a few basic requirements when interviewing a potential candidate:

  • Communications:  Is he an effective communicator?
  • Collaborative:  Can he effectively work across people and systems?
  • Dependable:  Can he add value to key processes by consistent actions with correlated research?
  • Teamwork:  Can he help in scaling, adding value and apply leverage to the entire team’s efforts?  Be part of the process to build and retain institutional memory and knowledge.
  • Presentation and Reporting:  Can he give progress reports and even self-evaluations to his manager?
  • Flexible:  Can he learn and adjust as events occur to improve?

After reviewing several of our successful customer implementations, and mapping to the “digital” employee requirements, I came up with a list of seven essential features to look for when selecting a virtual agent:

Understand end users’ requests accurately and have the skills to collect all requisite information:

The virtual agent must be able to fully understand the information he is working with, including the ability to process modular sets of skills, and acquire key information for different intents thru the chat, from the operational environment or from other systems. 

Inbound and outbound capabilities across people and systems:

Each employee contributes and participates in many business processes.  The digital employee must be able to not only take inbound conversations and events, but also to reach out to start conversing with end-users; take part in the process chain and/or originate workflows within itself.

Orchestrate, participate and optimising workflows with information:

A virtual agent should be capable of acting as an entry-level program manager for every task.  If given the right skills, the virtual agent can always do the grunt work of gathering information across people and systems to present a consolidated view for decisioning in each workflow.

Workflows are especially needed for enterprise service management (ESM) deployments.  The virtual agent must support response-driven as well as event-driven conditional branching and chaining.  An open integration is also a must for integration with a wide array of third-party software to seamlessly interact with minimal or no coding.

Execute actions that reduce the workload on support teams:

A virtual agent can reduce the amount of simple, repetitive tasks.  That is a win-win for everyone on the team. This includes:  answering common questions; initial creation and ticket submission; triggering fulfilment in external systems; using API calls; scheduling calls with agents and providing updates and following up post resolution to name a few.  These are all excellent ways to take advantage of a digital employee as a support assistant.  Furthermore, early detection is the best way to do less work.  A virtual agent can even help correlate crowd sourcing input from the end-users for early symptom detection such as a ransomware/malware scenario adding heighten visibility to the team.

Keep track of knowledge to record, update, and share at the right time:

A knowledge management capability is a must-have in a virtual agent.  Think of it as the long-term memory for the support and service organisation that can be shared.  This is essential for getting to the “self-service” scenario assisted by the virtual agent versus utilising a live member on the support team.

Insight feedback loop and KPI for fine-tuning:

The virtual agent should be able to autonomously record key metrics around use and interactions.  These KPIs need to be reviewed for targeting optimisation opportunities.

Acquire memories and act on future events

The virtual agent is the operating connection between incoming needs, deciding what the request is, matching it to what is known, serving up the knowledge (of course keeping track of what’s right, what needs updating and what’s no longer true), and acting on the knowledge through skills and process workflows.  The virtual agent must have the entry points for human or other inferencing elements on the various interactions, events and other active data sources as part of it’s daily operations.  This is a critical component.

Reduce support desk costs

The implementation of an intelligent agent is not just about streamlining and improving service, it also provides real cost savings.  Businesses can expect an overall support desk costs reduction of 30 per cent.  These savings are achieved in a variety of ways, including the following:

  • Ticket deflections—End users are more likely to embrace a self-service approach when they can naturally interact rather than filling out a form.   Virtual agents are also more capable then a typical service catalogue approach.  According to HDI, the cost of fulfilling a service request is $18-$175 and the cost of resolving a technical issue is $21-$110 in North America. When a virtual agent is available to ensure tickets are actionable, or deflected altogether, the cost improves to $1-3 in the ideal case when analyst does not need to be involved.
  • Eliminate repetitive efforts – Increase the support team’s available bandwidth by moving repetitive tasks to the virtual agent.   Think of it as having a support assistant for each of the support team members.  According to one survey, the attrition rates of level-1 agents are 30-45 per cent, around twice that of other business functions. 33 per cent of companies agree that improving the agent experience is a top priority. Less stressed and overworked agents have a lower turnover rate, meaning that companies don’t have to invest as much time and resources in recruiting, hiring, training, and onboarding new agents. According to the same study, the average cost of replacing an employee is 16 per cent of their annual salary.

The future of service management

From quantitative data to qualitative benefits, an intelligent chatbot or virtual agent will deliver a better service management experience and bring real business value.  Depending on the size of your organisation, the cost reduction can reach tens of thousands of dollars per month. The economic return of a virtual agent is huge and there has never been a better time to consider bringing one into your business.  This is the future of service management.

Wai, CEO, Serviceaide


Wai is Founder, President and Chief Executive Officer of Serviceaide ( A serial entrepreneur, he has been in the high-tech industry for 36 years. Wai held executive leadership roles at BEA where he was the EVP and Global Chief Product Officer and at CA Technologies where he was the SVP and GM of the Unicenter brand, SVP and GM of WW Services, and SVP and GM of interBiz.