A complete guide to conversational artificial intelligence.
- What is Conversational AI?
- How does Conversational AI work?
- Conversational AI vs. Chatbots
- Benefits of Conversational AI
- Conversational AI Use Cases
- Conversational AI by Industry
- Conversational AI Testimonials
- Implementing Conversational AI
In this article, we'll walk you through what conversational AI is, how it works, and why so many businesses are starting to adopt it. Whether you're interested in a conversational growth strategy or focused on conversational customer support, we'll break things down to set you up for success. You're probably familiar with some of the tools that use this technology, for example:
Each of these is a different application of conversational AI, among hundreds that have entered the market in the past decade.
What is Conversational AI?
Conversational AI combines natural language processing (NLP) with traditional software like chatbots, voice assistants, or an interactive voice recognition system to help customers through either a spoken or typed interface.
Two relevant supporting definitions:
- Machine Learning: This is the broad field of computer science dedicated to building software with algorithms that automatically improve themselves through repeated use.
- Natural Language Processing: This is a field of machine learning that focuses on modelling and interpreting language, such that software might effectively engage with humans.
How does Conversational AI work?
Conversational AI engages in contextual dialogue using NLP as well as other complementary algorithms. As one develops a larger corpus of user inputs, your AI becomes better at recognizing patterns and making predictions. Conversational Al engages with customers through four broad steps that we'll examine to get a better sense of this technology:
Step 1: Input Generation. Here, a user provides inputs either through voice or text.
Step 2: Input Analysis. If the input is text-based, natural language understanding (NLU) is applied to pull meaning out of the words provided. If the input is speech-based, ASR automatic speech recognition is first applied to parse sound into language tokens that can be analyzed.
Step 3: Dialogue Management. Here, natural language generation is used to create a response to a user's query.
Step 4: Reinforcement Learning. Here user inputs are analyzed to refine replies overtime to ensure their responses are correct and accurate.
Conversational AI vs. Traditional Chatbots
- Traditional chatbots often claim to have conversational capabilities, but humans have to write scripts and dialogues behind the scenes. The chatbot is told what to say in response to specific keywords. Your bot needs to be explicitly trained for every foreseeable scenario.
- A real AI chatbot conversation requires conversational AI, which does not need a script but rather progressively teaches itself through reinforcement learning.
Smart and Dynamic System
Deep Understanding of Customer
Sentiment Analysis + Data Insights
Not Scalable, Predictable Answers
Not intelligent, can lead to frustrating client experiences
Unable to meet customers where needed
Benefits of Conversational AI
Everyone from small-to-medium businesses to large enterprises can find great benefits from leveraging an AI platform. The benefits broadly fit into one of two categories:
- First and foremost, conversational AI helps deliver rapid responses to clients of all kinds. Customers are busy and impatient — giving them quick responses is one of the clearest ways you can make their experience more delightful.
- Following the initial response, conversational AI may also play an important role in educating users through helpful prompts and probing questions. It's incredibly common for customer support agents to give customers tutorials. With conversational AI, these tutorials can automatically leverage a client's profile data to ensure customers are receiving personalized guidance.
- Finally, conversational AI has proven itself a valuable guide for customers who aren't sure where to go. Within banks, for example, conversational AI is often instrumental in directing customers to the correct department when they call in.
- Starting at the beginning of your customer's experience, conversational AI can help you better convert website visitors into leads through intuitive, easy-to-engage with live chat windows.
- From here, conversational AI may help improve your overall conversion rate from lead to purchaser. By asking tailored questions, chatbots excite customers and support sales team efforts down the funnel.
- Finally, post-acquisition, conversational AI can play an instrumental role in triaging customer service inquiries. At Ada, we've seen this automated customer experience management divert as much as 80% of our partner's support inquiries, freeing up their human agents to focus on complex queries.
Conversational AI Use Cases
Use case #1: Customer Success Automation
- Online chatbots are rapidly replacing live chat agents. Conversational AI can answer frequently asked questions and offer personalized advice, often with a higher response-quality and a lower response-time than would be possible with humans.
- To this end, conversational AI often has a meaningful positive impact on both the bottom-line of your customer service function and your net promoter score. By creating extra time for human agents to problem solve the most-complex inquiries, customers end up with a far better experience at every part of their journey.
Use Case #2: Marketing Automation
Conversational AI changes traditional digital marketing in a few key ways.
- First, it makes onboarding effortless and intuitive by replacing predefined forms with user-driven messaging. Whenever you ask users to pick from a predefined list, there is a chance that your options are confusing. By instead allowing customers to write answers in their own words, companies are avoiding this confusion and improving onboarding completion.
- Second, conversational AI flips the concept of signing up on its head. Rather than hoping users will navigate to a specific page to express interest in your product, AI live chat allows them to do so from any page on your website.
- Finally, conversational AI is helping to triage leads more effectively. One way this happens is by combining lead enrichment data with user driven inputs to help direct leads towards the appropriate sales agent. Further, conversational interfaces allow for a great deal of “branching logic” without costly tech investment, allowing you to route leads in a more fine-grained manner.
Conversational AI by Industry
According to a Juniper Study study, the use of 'Conversational AI' or 'AI that you talk to' will grow to $556b by 2025 with a 125% growth YoY across the Insurance Industry. As digital and self-serve insurance brokers have continued to eat-up market share, legacy insurers have been actively looking for software that might improve their cost structure and reduce their customer churn. To this end, conversational AI has had a few breakthrough applications:
- Rapidly decreasing the time to claim resolution
- Reducing the hours of administrative work required from insurance agents
- Making consumer workflows more intuitive
- Leveraging natural language processing to help detect fraud
By automating workflows for repetitive tasks that staff have to undertake with conversational AI, incumbent insurers do a better job of retaining their client base.
Conversational AI in Financial Services & FinTech
Like insurers, financial services is an industry that's rapidly digitized over the past decade. As these organizations look to make their services more engaging, personalized, and retentive, they turned to conversational AI to deliver intelligent omnichannel experiences. Clients have reaped the rewards of this transition in several ways, including:
- Enjoying a broader the scope and scale of self-serve options
- Receiving greater personalization across one's banking experience
- Observing faster resolution times across SMS, web, and the phone
Here's one example use case for conversational AI in the financial services sector. A customer lands on their web banking portal and asks the live chat how to take out a loan. After being directed to resources on consumer loans, this same live chat application proactively asks the client if they would like to begin a loan application. After answering a few questions, the conversational assistant prompts our customer to finalize her application over the phone and shares the direct-line with an agent who can help her finalize the loan applications.
Conversational AI in Telecommunications
The IDC, a research institution focused on Artificial Intelligence, has indicated that around 63% of Telecom companies invest in conversational AI that helps optimize and improve their overall customer experience. Conversational bots serve an interesting role in this because they can serve the customer quickly and efficiently. Below are some of the key benefits these companies are enjoying:
- Far higher per-agent productivity across support functions
- Simplified escalation processes and workflows
- More robust data collection across their customer experience team
For all these reasons, the telecom industry has quickly applied conversational technology to eliminate repetitive tasks. From a customer delight perspective, telecoms have traditionally struggled with low NPS scores. This is why the improved measurement functionality inherent to conversational AI has been such a boon: continuous feedback allows the industry leaders to make data-driven decisions around how language impacts retention and CSAT. Further, some are now leveraging conversational AI to launch a/b tests evaluating how policy changes might better delight & retain customers. The result is a higher customer lifetime value, as well as an improved CSAT.
Conversational AI in Airlines & Travel
The travel industry suffered the brunt of the pandemic. A drop in utilization impacted both overall revenue and margins. Simultaneously, airline customer uncertainty has never been higher, with flight delays, border restrictions, and frequent cancellations. As a result, AirAsia's AVA team said they were handling 10x the normal number of queries: up to 500,000 per day in April. Their bot dealt with a 1,671% increase in support tickets coming through Facebook Messenger and WhatsApp, resolving 87% of cases. As a result, their conversational AI strategy helped:
- Radically reduce customer support costs
- Allocate >80% of their human agents to non-standard requests
- Improve customer service to mitigate churen during a tumultuous time
Airlines everywhere are seeing great success delegating refunds, flight change requests, and booking-detail inquiries to conversational AI. On average, each support call diverted saves airlines $2.20, which quickly adds up.
Conversational AI in Ecommerce
According to DigitalCommerce360, E-commerce sales have risen 44% from 2019 to the end of 2020. As adoption grows, so too do expectations from customers. Over 60% of eCommerce shoppers will leave a site if they can't quickly find what they're looking for or access help. For this reason, Juniper predicts that by 2023, 70% of all chatbots accessed will be retail-based. This statistic is in part a reflection of the widespread success retailers are starting to see leveraging conversational AI, mainly:
- Reducing their cost of acquisition by improving landing page conversion rates
- Increasing average order size the customer education and proactive recommendations
- Driving repeat purchases through proactive follow-up
Increasingly, conversational AI is driving improvements at every step of the funnel.
- Acquisition: When customers go inactive, live chat can help bring them back to the page. Further, by providing an intuitive on-ramp to purchase, conversational AI will often drive up landing page conversion rates.
- Activation: Many eCommerce retailers will utilize their chatbot to provide personalized shopping advice. Similarly, customer education delivered through conversational AI can expand average order size, making metrics at this step of the funnel more attractive.
- Retention: By collecting more data from customers through the purchase process and leveraging omnichannel communications, conversational AI can often drive up repeat purchase rates. Proactive touchpoints and personalized recommendations can keep customers coming back again and again.
Conversational AI in SaaS
According to Gartner, the SaaS industry is set to grow to $117.7B by 2021, achieving a 24% YoY. It's no surprise then that large SaaS businesses are getting to the point where they need to start implementing automated customer support solutions.
Our team at Ada has helped businesses like Square, Mailchimp, and many others radically improve their support functions' cost structure. If you're interested in finding out what we can do for you, sign up here for a demo.
Conversational AI Testimonials
"We've surfaced Ada to our customers so customers' simple questions can be answered immediately, and more complex ones are seamlessly escalated to a live agent. We've seen fewer tickets, faster resolution times, less stress for live agents, and most importantly, happier customers."
- David Clark, Senior Director of Operations & Analytics @ MailChimp
"[Conversational AI] is a better friend than your human friends. It's the only interaction that you can have that isn't judging you…a unique experience in the history of the universe."
- Phil Libin, Founder of Evernote
"Digitization and technology innovation play an increasingly big part in the banking sector, but as we all move faster than ever, we realize the need for meaningful human interaction - even with our bank. As we add new customers every day, we are scaling our operations quickly. We are achieving incredible customer satisfaction rates with the help of [conversational AI], exceeding anything ever seen before."
- Marco Puccioni, Head of Concierge & Partnerships at Buddybank.
Implementing Conversational AI
Curious if conversational AI is right for you? You try running the numbers with our Conversational AI ROI Calculator.
Of course, even if the numbers make sense, implementing conversational AI might take some time. Our team at Ada does our best to make this process easy for you. We have an entire team dedicated to implementations, where our focus is:
- Helping you conceptualize and then socialize your conversational AI strategy
- Training and launching your conversational AI to ensure its efficacy
- Providing recurring feedback and insights to your team, such that we understand what's working well and what can be improved
We've consistently seen companies achieve a 61 point increase in their NPS over just the first 30 days post-launch.
If you’re ready to get started now, you can sign up for our demo here.