पोस्ट

Successful Conversational AI: Blending Machine Learning & Human Intelligence, Part 2

By Chris Ezekiel, Founder & CEO

In February ISG, a leading global technology research and advisory firm, published their ‘ISG Provider Lens™ Intelligent Automation – Solutions & Services’ report. The report evaluates 19 conversational AI vendors against a set of market-driven criteria and places Creative Virtual firmly in the Leader category within the quadrant.

Recently I joined Mrinal Rai, Principal Analyst at ISG, and Jan Erik Aase, Partner and Global Head – ISG Provider Lens, for a discussion on conversational AI over Zoom. In my last post I shared Part 1 of our nearly half hour chat. During the first part of the discussion, Mrinal shared why ISG identified Creative Virtual as an industry leader in their report. Jan Erik and I also discussed current conversational AI trends as well as the evolving role of contact centre agents. You can watch Part 1 here.

In Part 2 of our discussion (scroll down to watch the video), Jan Erik and I address two more questions:

  • What are the biggest barriers organisations face when it comes to building, deploying, and maintaining successful projects?
  • What impact has the pandemic had on the implementation and usage of conversational AI tools?

One key barrier to success that we explore is not having a team with the right skills and experience. Often organisations try to tackle conversational AI projects internally with a lack of knowledge and a toolset that doesn’t enable them to scale the solution to different channels, additional departments, etc. or support enough users simultaneously as the project expands. This sets the whole project up for failure. When it comes to conversational AI, knowing what doesn’t work is just as important as knowing what does.

Jan Erik and I also touch on issues some projects face when integration points and APIs aren’t readily available or accessible. Creating a personalised, omnichannel support experience really needs the conversational AI tool to be properly integrated with other systems. The contact centre not being a part of the organisation’s digital strategy is another common barrier we encounter. This is starting to change, but until company structures are more joined up in this regard, we’ll continue to see this as a widespread challenge.

The need to have the contact centre as part of the digital strategy was highlighted over the past year by the pandemic. We saw record traffic to our virtual agents and chatbots in 2020 as customers turned to online self-service for quick answers to their questions. For many of the organisations we work with, having a well-established conversational AI solution was a lifesaver as their contact centres dealt with an overwhelming volume of calls at the same time as new public health measures designed to keep agents safe.

Having a human-in-the-loop combined with machine learning gave our customers the ability to change responses within their virtual agent quickly, safely, and securely so they could deliver reliable, up-to-date information. In fact, one of our customers found that updating their virtual agent was quicker and easier than updating content on their website. Their contact centre recognised that the virtual agent was helping to reduce call volumes and proactively provided feedback and new content to add based on incoming calls from customers.

Check out Part 2 of our ‘Successful Conversational AI: Blending Machine Learning & Human Intelligence’ discussion:

 

My next post will take a look at Part 3 of our session where we discuss setting goals and identifying KPIs for conversational AI projects. In the meantime, be sure to download your copy of the ISG Provider Lens™ – Conversational AI Quadrant Report.

Is Your Inexperienced Approach to Self-Service Driving Customers Away?

By Mandy Reed, Global Head of Marketing

As many people around the world reflect on more than a year of pandemic-related lockdowns, restrictions, and public health measures, it’s impossible not to marvel at how we all adapted to our current reality. This has involved a lot of trial-and-error as we learned new skills and created new routines in both our personal and professional lives.

Trial-and-error is important to our individual growth and life in general. However, when it comes to areas like customer support, using a trial-and-error approach can have a devastating effect on your customer experience. And it is no secret that poor experiences can lead directly to customer churn and lost revenue.

Perhaps in the early days of limited in-person interactions and surges in calls to contact centres, customers were more understanding about long wait times or out-of-date self-help content. The attitude of ‘we’re all in this together’ extended to giving companies a little space to try some trial-and-error to get their support experience right. If that grace period did exist, it is now long over!

During the past year, more customers have turned to digital channels and automated self-service for support. Usage of virtual agents and chatbots exploded with record-breaking levels of traffic. Customers expect these conversational AI tools to be easy-to-use, convenient, accurate, and reliable. When built and maintained properly, they are all those things.

Unfortunately, not all companies tackle the implementation of conversational AI solutions with a realistic understanding of what it takes to make them successful. Instead, they take a DIY approach with limited internal knowledge and experience. This requires a lot of trial-and-error which creates poor performing tools and frustrated, unhappy customers.

As Claudio Chico, Development & Support Technician at Creative Virtual, explained in a recent conversational AI guide:

“A proper business virtual agent has many parts and building one involves knowledge in many areas. If any part is new to you or you aren’t extremely familiar with the tools you are using, you’re stuck applying the principles of ‘hoping this works’ and ‘changing stuff and seeing what happens’. When you outsource this to an experienced provider, you get a whole team of people who not only know what they are doing but have done it thousands of times. They have a deep understanding of how to use their tools and how they work, so nothing is a mystery anymore.”

Part of the underestimation of the importance experience plays in successful conversational AI projects stems from a misconception that chatbots and virtual agents are new self-service solutions that burst onto the scene several years ago. If this is new technology, then surely everyone is inexperienced and utilising a trial-and-error methodology, right? The truth is this technology has been used in areas such as website self-service for over two decades.

Even though these self-service tools may be new to your company and team, vendors like Creative Virtual have years and years of experience with delivering successful solutions. This means that forward-thinking companies – perhaps even some of your competitors – have years and years of experience with offering successful solutions. It also means that customers have used those successful solutions when engaging with other businesses and will use your chatbot or virtual agent expecting the same level of reliable and accurate self-service.

When it comes to creating positive customer experiences and getting the most from conversational AI technology, there is no substitute for having hands-on experience with building, integrating, installing, maintaining, and expanding virtual agents and chatbots. An inexperienced, trial-and-error approach doesn’t drive success. It drives your customers away.

Download the Guide to Selecting a Virtual Agent or Chatbot Vendor: Forget the Technology & Focus on Experience whitepaper for more tips from industry experts.

Also check out the ISG Provider Lens™ Intelligent Automation – Solutions & Services report for the analyst group’s independent evaluation of the conversational AI market and vendors.

Bottom line: Work with an experienced team to deliver your company’s self-service solutions and leave the trial-and-error for finding the most flattering lighting for your next Zoom meeting or testing the best ways to trick your kids into eating their vegetables.