The Chatbot 2.0 – An Introduction to the Puzzel Smart Chatbot

Charlie Mitchell

Discover how AI advances are powering the next generation of chatbots

Insights
The Chatbot 2.0 – An Introduction to the Puzzel Smart Chatbot

In February 2019, Forrester Research revealed that 54% of US online consumers believe conversing with a chatbot will have a negative impact on their quality of life.

Fortunately, the industry has made great strides since. Very few models now operate with only natural language understanding (NLU) and scripts. Advances in machine learning, natural language processing (NLP), and robot process automation (RPA) are transforming bot technologies into something much more powerful.

Enter the Puzzel Smart Chatbot. Alongside answering all those simple, transactional customer queries much more accurately, it assists customers with various other critical tasks to CX success. These include:

  • Website Navigation – It holds the customer’s hand through various online processes, sharing additional information when they appear stuck. In addition, if the customer types in a query into the website search engine, the chatbot takes them to the most appropriate page while highlighting critical pieces of information that relate to the question.
  • Form Filling – The bot can present forms within the chat window, automating particular inputs by retrieving and verifying customer information with a CRM system.
  • Automated Check-Ins – Linking with various data sources, customers can check the status of balances or processes through the chatbot. Delivery tracking is an excellent example, made possible with an integration to the order management system.
  • Built-In Escalations – If a query is too complex, the chatbot can transfer the customer to a live contact centre agent, pulling with it the transcript. As a result, the agent immediately understands customer intent and can continue the conversation without the customer having to repeat themselves.

As referenced, many of these benefits stem from enterprise integrations, particularly with the Puzzel Contact Centre, which enhances the capabilities of the Smart Chatbot beyond many similar models within the bot space.

How Else Does It Surpass Other Chatbots?

Ram Sagoo, Sales Director at Puzzel, identifies the following three differentiators of the Puzzel Smart Chatbot.

  1. Redeployments Across Other Channels

Not only can the chatbot reduce incoming call volumes through automating online support, but  Puzzel offers a service to redeploy successful text-based chatbots as voicebots.

Thanks to its leading automatic speech recognition and RPA capabilities, customers can get the information and help they need quickly, 24/7, no matter the channel.

Through its Bot Gateway, Puzzel can also connect to third-party chatbots – such as those offered by IBM and Dialogflow – bringing them to life across various other engagement channels too.

  1. Automating Agent Tasks to Streamline Service Experiences

Unlike many basic bots that replace FAQ pages, the Puzzel Smart Chatbot handles basic tasks that agents typically deal with. These include:

  • Helping customers open an account
  • Assisting customers in logging into their accounts
  • Sending information about an invoice
  • Going through security checks before passing over to human support

The bot can perform these tasks while guiding customers through other online processes step-by-step and fetching data from various enterprise systems. As such, it complements online self-service portals, driving down bounce rates to the contact centre.

Also, by implementing this use case, companies can increase the uptake of such self-service channels, with existing customers achieving a 50% cost reduction per service request, on average.

  1. Proactive Web Services

Combining the chatbot with the Puzzel Digital Engagement solution paves the way for companies to offer proactive services on their website and enhance online experiences.

To do so, it creates a profile for all website visitors, not only existing customers. As a result, everyone receives a personalised, proactive service based on their digital behaviour.

In addition, companies can connect digital services, eCommerce, and customer services to deliver a more satisfying digital journeyw, from the first visit to purchase and post-sales support.

Finally, Puzzel Digital Engagement gathers digital behavioural data, combines it with other customer data to predict intent, and guides the customer towards the optimal route of resolution.

Who Is Implementing the Puzzel Smart Chatbot?

Many B2C companies across the financial services, energy and utilities, telecommunications, and retail sector have implemented the Puzzel Smart Chatbot.

In doing so, most have opted for an omnichannel approach, adding bots to both traditional and new digital channels, catering for a diverse customer base.

Such a process is easier for companies that have transitioned to CCaaS software. Yet, Puzzel can also support the migration process, enabling integrations that enhance bot capabilities and more innovative service experiences.

In terms of the Puzzel Smart Chatbot alone, it offers the most value to businesses with an operating model that:

  • Complements multi-skilled agents and extended customer service hours
  • Balances cost-efficiency with a drive to improve CX
  • Wishes to closely monitor the performance of automated systems
  • Fixates on simplifying the agent experience through bot-driven automation

As with any chatbot implementation project, a small IT team is necessary to bring particular capabilities to life and manage the vendor relationship – especially when installing multiple bots.

A Puzzel Smart Chatbot Case Study

ESCI is an American student loan servicer that partners with 1,500 colleges and universities across the country, providing services such as:

  • Past-due accounts receivable management
  • Campus-based student loan servicing
  • Tax document services
  • Tuition payment plans
  • Refund management
  • Call centre outsourcing services

Serving nine million students nationwide, the ESCI answers 1.3 million calls and chats every year.

The Challenges

On average, ESCI receives around 1,000-2,000 chats a week. Yet, this can spike to almost 11,000 a week during the tax season, making demand management a significant concern.

During these peak periods, staffing the contact centre is tricky, especially given recent years of problematic recruitment climate.

Fortunately, most of the incoming contacts were repetitive, routine enquiries through the live chat channel, popular with young students.

With this in mind, the ESCI implemented Puzzel’s Digital Engagement and Smart Chatbot solutions to handle this demand in the most efficient, customer-friendly way.

The Solution

The ESCI chatbot answers tax, student loan, and financial assistance queries while also handling routing tasks for students that include:

  • Finding out when payments are due
  • Resetting passwords
  • Resending emails
  • Completing forms
  • Locating specific information on the website
  • Supporting navigation of online applications

In addition, the bot streamlines identification and verification (I&V) processes before each contact begins. Then, it guides students to a menu where they can select the topic of their query and receive personalised, automated customer service.

According to Sagoo, such service includes guided journeys, where students choose between various answers and free text journeys that allow them to type questions and responses. He stated:

“Free text journeys provide an excellent method for capturing the voice of the customer (VoC). Alternatively, guided journeys and forms help collect and verify information with less chance of error.”

Either way, if the query is too complex for the bot to answer or the customer does not have all the necessary information – i.e. they do not remember their student number – they automatically transfer to a live agent.

The Results

The ESCI Smart Chatbot now handles 75% of all chat interactions, with only 25-30% transferred to an agent because they are overly complex.

In addition, the average handling time (AHT) has fallen to nine minutes, saving the ESCI 390 hours per month, with the bot being able to handle three queries concurrently per customer.

Notably, the bot has also significantly lowered staffing requirements during the peak tax season, reducing costs further and simplifying onboarding, coaching, and HR stresses.

What Does the Future of Chatbots Look Like?

The Puzzel Smart Chatbot will continue to evolve in many ways. For instance, it may soon incorporate image recognition to deal with a broader range of customer queries.

Gesture recognition is perhaps another capability to build into these bots, which may automate video interactions through smart displays – the next evolution of smart speakers – in the future.

Perhaps slightly closer to the present day, however, is the dawn of the proactive voicebot. Such a bot sends outbound updates to customers through a neural voice that performs AI-based voice cloning. Sharing an example of how this could work, Sagoo said:

“Instead of a customer contacting the company about a delivery, the voicebot contacts them proactively. This could be to confirm if you’re going to be in, or if you wanted to make a change to your delivery.”

However, there are many scenarios and verticals in which a proactive voicebot can free agents from repetitive outbound tasks and complete various processes.

Whatever the case, the voicebot will still transfer the customer if necessary or transfer a smartphone user to an online conversation, if they prefer.

 

Eager to learn more about how the Puzzel Smart Chatbot is enhancing digital CX? Then, head on over to the Puzzel website to dig deeper.

 


Join our Weekly Newsletter