Even before the pandemic caused long wait times to reach an agent, there was growing consumer interest in using self-service to resolve questions and issues. Now, a Gartner survey conducted in 2021 shows that 71% of customers prefer to resolve their issues without having to contact customer service1.
Self-service can be a win-win for companies and their customers. Effortless self-service experiences increase digital adoption and boost customer satisfaction and Net Promoter Scores, while deflecting interactions from contact center agents.
Yet, many companies have failed to deliver on the benefits of self-service. Customers still find it difficult and time consuming to get helpful, relevant answers across existing web self-service, knowledge management and web content management solutions. In fact, a Gartner survey in 2020 reported that only 13% of customers are fully contained in self-service2. That means customers must often start over in another channel to get the assistance they need, creating an even more frustrating experience.
Given the past failures in self-service and current contact center challenges, it’s more important now than ever for customer experience and contact center leaders to embrace conversational artificial intelligence (AI) for human-like, self-service interactions. With an intuitive, conversational experience, companies can improve self-service completion rates, enhance the customer experience and reduce contact center volumes and cost.
This guide can help you understand the features and capabilities you should be
looking for in a conversational self-service solution.
What you’ll learn in this guide
The difference between simple chatbots and intelligent virtual assistants
How to create the business case for self-service
Whether it makes sense to build an intelligent virtual assistant yourself
Essential capabilities in a conversational self-service solution
Why a conversational automation platform is important
The difference between Traditional Chatbots and an Intelligent Virtual Assistant
Any discussion about conversational self-service needs to begin by addressing the questions around chatbots and their relative success or failure for self-service.
Traditional chatbots are typically simple, rule-based self-service tools that follow a predefined workflow and are tightly scripted. Usually text-based only, the primary use case is to help customers get quick answers to frequently asked questions without having to speak to an agent.
As a first-generation technology, chatbots had a role to play in introducing customers to guided self-service. However, they inherently have drawbacks and limitations in terms of their functionality and user friendliness. And chatbots are typically siloed from the rest of the customer experience, so if customers switch channels to get the help they need, they are forced to start their journey over again.
Fast forward nearly a decade. Today’s intelligent virtual assistants (IVAs) built using advanced conversational AI are in a different class altogether than traditional chatbots. A conversational IVA can deliver a human-like experience, simulating human conversations and delivering frictionless experiences in multiple channels, including text and voice. Multimodal capabilities allow an IVA to combine inputs via text, voice and touch (on a screen) with the ability to process data from various additional sources, such as images, videos, tables, maps and more.
All of this means that conversational IVAs can assist customers with far more than getting answers to FAQs, and they can do so in a user-friendly, human-like way. These tools can engage customers for routine journeys across their channel of choice, helping them with requests such as:
- Setting up a new bank account
- Booking a trip
- Submitting an installation request
- Scheduling an appointment with a healthcare
IVAs can also automatically and seamlessly transfer higher-need inquiries — together with context, intent and other AI-obtained information — to agents for more personalized care and handling.
Making the Business Case for Conversational Self-Service
In fact, with the right conversational IVA, contact center leaders can achieve measurable, sustainable improvements in:
- Self-service completion rates
- Customer satisfaction (CSAT) and NPS
- Scalability to handle higher volumes of customer interactions
- Operational costs in the contact center
- Reducing friction in customer journeys
To begin creating your business case for investing in a conversational IVA, consider the following use cases, questions to ask to uncover specific pain points and related metrics, and the benefits your company can anticipate.
IVA Use Case: Deflecting calls from agents to self service
Questions to ask
- How much of your contact center volume could be resolved with conversational self-service?
- If you already have self-service capabilities, what is the completion rate?
- How often do customers have to switch to a new channel to resolve their issue?
- How often do customers ignore your self-service and pick up the phone instead because of previous poor experiences?
IVA Benefits
- Reduce volume of contact center interactions
- Reduce costs by deflecting to lowercost self-service channel
- Improve completion/containment rates in self-service
- Improve scalability to handle spikes in volume
- Extend customer support to 24x7 without additional staffing costs
IVA Use Case: Increasing revenue and customer lifetime value
Questions to ask
- Are you missing opportunities in your current selfservice to cross-sell and upsell?
- Can your self-service capability personalize recommendations for upselling/cross-selling based on understanding of customer context, intent, emotion and sentiment?
- Do you have the ability to track the customer journey and proactively message across all channels?
IVA Benefits
- Take full advantage of opportunities for upselling and cross-selling
- Increase revenue
- Grow customer lifetime value
IVA Use Case: Improving customer satisfaction
Questions to ask
- How easy, fast and convenient is your current self-service functionality?
- Do you offer customers multimodal self-service with consistent experience across web, mobile and voice?
- How often do customers have to switch to a new channel to resolve their issue?
- How long are wait times for customers to reach an agent?
IVA Benefits
- Decrease customer effort and improve self-service convenience
- Deliver consistent omnichannel experiences
- Reduce customer wait times for an agent
- Improve NPS scores and CSAT
IVA Use Case: Delivering an omnichannel experience
Questions to ask
- Does your current self-service capability offer the same experience across multiple channels?
- What’s the current cost and effort involved to maintain and update your self-service/chatbot capabilities across channels?
IVA Benefits
- Optimize customer journeys and remove points of friction
- Deliver consistent omnichannel experiences across web, mobile and voice/IVR
- Reduce your company’s effort to optimize self-service outcomes
- Eliminate silos and the additional overhead of maintaining separate experiences in different channels
IVA Use Case: Deploying intelligent call routing
Questions to ask
- When customers need agent assistance, are they forced to start over, or does context automatically transfer to the agent?
- How much time do agents spend gathering and confirming information that was already identified in the self-service experience?
IVA Benefits
- Improve agent and customer experience through seamless escalations
- Reduce average handle time (AHT)
- Increase first call resolution (FCR)
- Improve NPS and CSAT scores
Can you create your own IVA in house?
With the availability of open source chatbot platforms and conversational AI toolkits, your company may be considering taking a do-it-yourself approach to automating self-service. While it’s possible to build your own traditional chatbot, most companies will struggle to create a fully functional conversational IVA using open source or AI frameworks.
Choosing an off-the-shelf conversational IVA solution that is customizable by non-technical business users will almost always be faster, more cost effective and deliver greater return on investment than taking a DIY approach.
That’s because building and integrating a conversational IVA capability from scratch demands technical expertise and experience, as well as an ongoing development program and resources to optimize machine learning and implement new functionality. Your organization may have some or all of the technical resources required in house; but before you decide, consider the following:
Scope: Will you be able to create a conversational IVA that handles all the customer journeys you want to support via self-service? Which channels will you need to support with your IVA?
Timing: How long will the project take? How long are you willing to wait before the IVA is fully functioning and ready to use?
Resources: How many engineering resources and conversational AI experts will you need? Will you need to hire consultants?
Integration: How many systems will your IVA need to integrate with and how will you handle the integrations? Do you have the resources to maintain custom integrations as your contact center and AI-based systems evolve?
Budget: Based on scope, resources and integrations, what will be the total cost of the project?
Maintenance: What will it cost to scale and maintain the IVA as new capabilities are introduced or new customer journeys added?
Evaluating a Conversational Self-Service Solution
Once you’ve established the use cases and business case for conversational self-service, it’s time to search for the right solution. To deliver a human-like, conversational self-service experience, you’ll need to look for an IVA that offers the following core technologies and capabilities.
Domain-specific, conversational AI
Domain-specific conversational AI has made huge leaps in sophistication by focusing on narrow applications, such as customer service, in combination with specific industries, such as financial services, telecommunications, healthcare, travel & hospitality and others. By looking for a conversational IVA solution that is specific to your industry and to customer service, your company can optimize its understanding of customer conversations.
Industry-leading speech and text recognition
Your conversational IVA needs to understand your customers with a high degree of accuracy. To do so, it must be able to listen for and automatically detect for language, including specific dialects for voice modalities. It should also support the languages used by your current customers as well as your future language needs as your company expands into new markets.


Intent recognition and analysis
A conversational IVA solution should be able to identify and understand what the customer wants to achieve, using machine learning to continuously tune and optimize algorithms to deliver the highest degree of accuracy possible in predicting true intent.
Customer sentiment and emotion recognition and analysis
Another important capability is sentiment recognition, which provides insight into the customer’s state of mind. At a minimum, your conversational IVA should recognize, extract and score customer sentiment as positive, neutral or negative. However, to better understand your customers’ feelings within the context and intent of the interaction and take relevant actions based on that understanding, you should choose a solution that also identifies customer emotions such as sadness, frustration, anger and happiness.
Advanced dialog manager and natural language generation
For conversational self-service, an IVA must be able to generate a response in natural language for a human-like interaction. This includes the ability to recognize when a customer switches context and to shift the conversation to support the new topic or intent.
Intelligent call routing/Human-in-the-loop design
Critical for self-service experience management, your IVA needs to make the transition to a human agent effortless for both the customer and agent. It should intelligently recognize the customer’s intent, route the call to the best live agent to support the customer’s query, and transfer the call with full context so the conversation remains seamless across channels.
No-code designer and low-code integration
To deliver consistent experiences across channels and reduce the effort and overhead of adding new customer journeys and automation, look for a solution that makes it easy for business users to create, customize, monitor and update the conversational IVA. The solution should enable business users to design once and reuse the journey across multiple modalities, channels and touchpoints. It should also support programming for upselling and cross-selling based on intent, emotion and sentiment. Likewise, low-code integration enables you to connect the IVA with disparate systems to orchestrate and automate the customer experience across touchpoints.


Maximizing Business Value with a Complete Conversational AI Platform
Applying real-time conversational AI across the entire customer journey — not just self-service — can help your company realize far greater returns and strategic business outcomes than with siloed solutions. As part of your evaluation of conversational self-service solutions, consider whether the IVA is part of a broader platform of AI-powered capabilities.
With a platform approach to conversational AI, your company can create the operational backbone for driving contact center transformation. For best-in-class transformational results, look for an enterprise-grade conversational AI and automation platform that offers the following capabilities:
-
Attended and unattended robotic process automation (RPA):
Conversational AI platforms that incorporate RPA capabilities enable contact centers to automate manual agent tasks, such as updating customer relationship management (CRM) systems, sending followup emails to customers and more. Sophisticated platforms include the ability to recognize promises (commitments) made by agents and to automatically create and perform follow-up tasks. -
Intelligent applications:
AI-powered software called intelligent applications handle specific use cases within the contact center, such as intelligent agent assistance. A contact center-specific conversational AI platform with intelligent applications enables you to optimize and automate the end-to-end customer journey. -
Interaction analytics:
Intelligent decision support uses machine learning and reasoning to discover insights, find patterns and uncover relationships in data, automating the steps that would take humans days—or longer—to exhaustively analyze all the conversations across contact center interactions with customers. -
Voice biometrics and security:
This capability applies advanced conversational AI to agent verification, using ongoing voiceprints to authenticate agents continuously during their shifts. This is an important feature to reduce contact center fraud and build customer trust. -
Reporting and dashboards:
Built-in reporting provides insight into conversational performance and the customer journey. The best solutions offer custom dashboards and visualizations, text search, topic discovery and collaboration via shared workbooks. -
Security and compliance:
Depending on your industry, look for a solution that supports the regulations that impact your company, such as PCI-DSS, HIPAA and GDPR. In any case, the platform and the interaction analytics solution should support data redaction of personally identifiable information (PII). -
Enterprise-class/scalability:
To understand and optimize 100% of your customer conversations, you need a scalable, enterprise-grade platform built to take advantage of native cloud elasticity. It’s also important to make sure the platform includes an API gateway to enable connection to data as well as the ability to ingest insights and sensors from external systems.
Choosing the Right Provider
While evaluating the technology and capabilities you need to support your conversational self-service needs is vital, choosing the right provider can ensure your program’s success. The ideal partner should offer:
- Visionary leadership and a technology roadmap that aligns with your vision
- Deep expertise in AI, NLP/NLU and related technologies
- Domain expertise in optimizing customer experience/ contact center operations
- Deployment methodology and services to help you achieve rapid time to value
- Security and privacy of customer data
Next steps
Effortless self-service experiences are essential for improving customer satisfaction and Net Promoter Scores, while deflecting interactions from the contact center. IVAs built on a conversational AI platform deliver a human-like, easy-to-use and multimodal
self-service interaction that improves completion rates, enhances the customer experience and reduces contact center volumes and cost.
Choosing the right solution starts with defining the use cases for your company, creating a solid business case, understanding the capabilities you will need and carefully evaluating your options. We’re here to help as you take the next step in your conversational self-service journey.
1. “Younger Generations Increasingly Self-Serve Through Non-Company Digital Channels,” Deb Alvord, Smart Customer Service, May 2021
2. “Enable Customer Self-Service to Deliver Better Support,” Gartner, 2020