Wynston Loading Icon

Feb 14, 2019 | By Matt | Share this article


There are two types of Conversational AI, and distinguishing between the two can be troublesome for any organization.

The ability for your customers to contact you and converse with an AI before automatically connecting them with a live person is now considered a best practice in B2C companies.  This wasn’t the case just a few years ago, but the benefits have become crystal clear.  They include:

  • A fast response, any time of day, every day of the year
  • The ability to connect through many channels, such as Facebook, WhatsApp, Webchat, Alexa, etc.
  • A consistent answer regardless of channel
  • Higher conversion rates and lower care costs
  • Customer’s expect this option to start, even if they end up with live support

As this market emerged, many vendors piled in to offer a virtual assistant toolkit to allow companies to build their own.  This is typical in the early stages of a new technology cycle when it is immature.  Vendors initially provide tools, but not solutions.  Many brands rush to test the new technology, and then reality sets in.

As a new technology matures, the buyers realize they’re looking for solutions, not technology.  They want to solve a problem, not deploy technology.  And this is where White Glove Conversational AI comes in.  Most brands who have tested an AI-based virtual assistant have realized that the challenge isn’t to find the smartest AI, but to find the AI that will most effectively learn.

There are two types of conversational AI, White Glove and Do It Yourself (DIY), and it’s important to understand the difference between them.

White Glove Conversational AI is:

  • A full-service conversational AI solution
  • A Virtual Assistant technology stack
  • A suite of AI optimization tools
  • An experienced AI management team
  • An extensive library of AI assets
  • It delivers a solution to your customers that learns every day
  • It delivers insightful Data about your customer behaviour that supports business decision making
  • It comes with a guaranteed level of system performance

DIY Conversational AI is:

  • A Virtual Assistant technology stack

Most companies jumped into DIY Conversational AI to see what it could do for them.  At this stage, brands were testing the technology, and toolkits work great for a test.  Unfortunately, when the time came to roll the Virtual Assistant out to millions of real customers, they realized the operational requirements go far beyond toolkit features.

Over the past few decades, many technology investments haven’t produced the promised value with millions spent and little to show for it.  This is usually due to a lack of ongoing attention or full understanding of how to operate the technology within an organization.  The migration to cloud is a great example of this trend.  Thousands of companies came to the realization that they weren’t experts in operating data centres (DIY) and that it was much more efficient to purchase a service (Cloud) than a pile of hardware and operate their own.  Most large companies are now realizing that the attention and tools required to make a VA succeed are outside of their domain of expertise as well.

I’m not suggesting that a DIY approach won’t work for some companies, just that they should go into the decision making process with all of the information.

Here are some guidelines for choosing between white glove and DIY conversational AI:

1. Do you have the AI management skills in your organization and the ability to keep them focused on your virtual assistant training?  Do you have linguists, machine learning experts, conversational designers, analysts and content managers who can dedicate a large portion of their time to ensuring the virtual assistant is improving every day.  If you have an effective team then DIY may be the better choice, if not, then a white glove service will deliver the experienced experts you need when you need them.

2. What is the volume of annual customer transactions?  Anything below 1 million interactions may be better suited to a DIY approach.  If more, you’re probably going to see more value with a white glove approach.

3. Do you have a management team that will continue to support and fund both learning and new features for your virtual assistant? This technology is evolving fast and will require attention and budgets for many years to come.  A lack of support equals failure and bad customer experience.  If you can’t guarantee exec attention (and budget) then white glove with its predictable cost structure will be a better choice.

4. What is the level of complexity of your customer interactions?  If you need to build a virtual assistant to take pizza orders or hail a ride, then DIY may be the better choice given the simplicity and limited use cases you’re facing.  If your customers require more complex interactions such as sales engagement, billing and service questions, or multiple language support across various lines of business, then a fully managed white glove service will deliver a better experience for your customers.

The Conversational AI market is still in its infancy, but it is evolving at an amazing rate.  The market is now going through the normal transition from technology curiosity to a solution that delivers real value to customers.  At this stage, new technology starts to separate into two categories, technology and solution.  I hope this helps you choose the type of Conversational AI that is right for your organization.