By Leslie Swanson, Founder and CEO of eXalt Solutions
Chat Bots in the Past — The Turing Test
In 1950, Alan Turing published “Computing Machinery and Intelligence” and proposed the Turing Test –
“The ability of a computer program to impersonate a human in a real-time written conversation with a human judge, sufficiently well that the judge is unable to distinguish reliably — on the basis of the conversational content alone — between the program and a real human”
The first Chat Bots were designed in the 1960’s. Artificial Intelligence had many false starts in the 1980’s but due to computing limitations it never made it into the mainstream.
Chat Bots of Today
Gartner estimates that by 2020, 85% of customer interactions will be without a human agent!
This prediction is clearly backed by customer preferences:
· 75% of B2B prospects prefer to buy online
· 70% of people prefer texting over calling
· In 2015 Gartner forecasted that application usage was beginning to plateau
· According to one report, next year 20% of firms will look to add voice enabled interfaces to their existing point-and-click dashboards and systems.
· A recent study from Gartner suggests that the percentage of customer service experiences that will be handled by conversational agents is expected to increase from 3 percent in 2017 to 30 percent by 2022
· Messaging apps have surpassed social networks (see graphic below)
Chat Bots and conversational interfaces are simple and not to be underestimated in terms of their power.
· No Learning Curve: There is no new UI to learn and you don’t have to READ FIRST. In contrast, traditional web application User Interfaces are difficult: You have to find the app, sign in, learn the UI, and process potentially hundreds of inputs on the screen before you make a selection. With a Chat Bot there is no learning curve and no confusion as to where to click.
· Instantaneous Response Times: With a Chat Bot there are no web loading times. A user can take action immediately.
· Unified Interface: Chat Bots can take actions that you may have to do manually on multiple websites.
Chat Bots are ideal for sales
One of the major investment areas for AI technology will be in sales. And Chat Bots are the key AI technology that have been addressing people-less customer interactions.
But many in B2B have been dismissing Chat Bots as a B2C marketing fad that handle simple requests, with predictable dialog such as customer service, but that are far too limited for sales.
Why Chat Bots have failed to get traction in B2B sales
The only reason that Chat Bots have not been applied as widely to B2B sales is the level of “Knowledge” that is required to drive a meaningful conversation. Here is a quick checklist of what is needed for Chat Bots to assist in B2B sales transactions:
Key Requirements for a B2B Chat Bot — Knowledge
· Ability to Address Relevant Use Cases
· Ability to build a Chat Bot for complex conversations
· Ability to keep Context or memory of prior input and modify downstream responses
· Ability to transfer to the web and back
· Ability to transfer to right human resource(s) at the right time
· Ability to communicate with existing business systems
For the reasons above we believe what is needed is a Knowledge Bot not a Chat Bot.
1. Ability to Address Relevant Use Cases
B2B has several Chat Bot use cases but the most compelling for sales is Lead Generation through advisement.
For Lead Generation many companies are still relying solely on phone and email. To gather leads companies place a blog or white paper with a call to action form. When the user:
· Reads the White paper it doesn’t advise them of a personalized solution to meet their specific needs.
· Completes the Form they are told “we will get back to you”.
Now imagine providing a Chat Bot to talk to those leads in real-time? Imagine the Chat Bot having the intelligence of your best experts with the ability to:
· Ask qualifying questions
· Provide solution recommendations
· Offer seamless transfer to a person or a system to complete a transaction.
Suddenly the Chat Bot can have a significant impact on sales!
Knowledge Bots have the intelligence of your best experts. They can advise and be full-fledged recommendation advisors as well as perform rules based prospect qualification.
2. Ability to build a Chat Bot for complex conversations
What has stopped companies from offering a Chat Bot for advisement and qualification is the complexity of the conversation. Most Chat Bots are similar to old fashioned IVR systems for the phone — a tree-based static dialog. But B2B requires more than just a simple decision tree for dialog. Knowledge Bots use a “codeless” Knowledge Base for Deployment: With one instance of rules or meta data put into a Knowledge Base platform, users can deploy a web form experience and/or a Chat Bot experience with almost unlimited levels of dialog.
Knowledge Bots are generated from a Codeless Knowledge Base and remove the manual effort associated with generating hierarchical dialog. With one instance of rules or meta data put into a Knowledge Work-as-a-Service platform you can generate the web form experience or a Chat Bot experience that is more flexible than a simple hierarchical dialog.
3. Ability to keep Context or Memory of prior input and modify downstream responses
Typically today’s Chat Bots can’t hold contextual information for longer than a few chat bubbles, and will end up losing track of what the user was saying before they posed the question.
Chat Bots can have a conversation but the possible responses to any question are fixed. Knowledge Bots can adapt the conversation and present variable choices based on prior parts of the dialog. Notice that the Knowledge Bot presents different choices based on historical answers:
Knowledge Bots are able to conduct dynamic conversations — they remember prior parts of the conversation and are able to adjust downstream responses.
4. Ability to transfer to the web and back
In the two Interactions above you can see that the Knowledge Bot handed the user off to a link on the website.
In B2B the Knowledge Bot must continue to help the user through the interaction!
Users require Knowledge Bots that are part of a seamless session that is shared across both the web and chat!
Knowledge Bots allow users to seamlessly transfer between web and chat.
5. Ability to transfer to right human resource(s) at the right time
Unlike a Chat Bot, a Knowledge Bot utilizes a full rules engine to help assist real-time collaboration. Many B2B sales conversations require the participation of different domain experts. For example, a Server Solution may require a Server Engineer, Storage engineer and a Software specialist in licensing. If a potential buyer would like to have each of these experts review the proposed solution and answer some questions they want to have the Knowledge Bot ROUTE one or many requests for them.
Knowledge Bots facilitate real-time collaboration on the web and/or chat with the optimal resources.
6. Ability to communicate with existing business systems
If a customer has been empowered to determine the optimal solution for their business problem they won’t need to speak to an agent! They will want to convert to purchase. This is where Chat Bots are performing well today.
But in B2B purchase often involves an entire ecosystem of business partner systems at distributors, vendors or resellers who sales reps have to interact with to determine the price for a user. Promotions and availability are often determined through a chain of phone calls and email requests. These requests often consume 70% of a sales rep’s time! This is where a deeper rules-based Knowledge Bot can excel.
Knowledge Bots are a unified user interface for other websites and systems — taking independent rule based actions and effectively merging systems and clicks across email, websites, messaging systems into one interface.
Chat Bots are being viewed as one of the best methods to engage customers in B2C by offering instant response times in an easy to use conversational interface. However, these traditional chatbots are not equipped to address the complex requirements of today’s B2B customers who expect in depth answers to their questions. To address this, companies need to incorporate the use of KNOWLEDGE BOTS, not CHAT BOTS into their customer experiences. A Knowledge Bot is capable of adapting the dialog for complex conversations based on knowledge-based Rules more extensively than a Chat Bot.
More importantly, today’s B2B sales teams must have a “Better Together” strategy with Knowledge Bots assisting sales reps throughout the sales process. Reps can use Knowledge Bots to qualify prospects, increase lead generation, collaborate with partners, offload administrative tasks and provide users with quotes.