4 KPIs For Measuring Conversational AI And Chatbots In Customer Service Automation

Post by
Indrek Vainu
March 29, 2021
4 KPIs For Measuring Conversational AI And Chatbots In Customer Service Automation

1. Self-Service Resolution Rate


Self-service rate is also known as the deflection rate. It measures the number of users that had a conversation with the AI and did not continue chatting with customer service agents as a percentage from the total number of users that had a conversation with the AI. 

This metric shows the % of users that otherwise would have chatted with your customer support agents but now chatted only with the AI. This is the amount of chats that the AI handled on its own and that your customer service agents did not have to do.


A good general indication of the impact that customer service automation platform can have on your customer support. A great customer service KPI for automation. Shows a clearly measurable automation rate of the smart AI chatbot.

Average benchmark


2. Cost Saved With Conversational AI


Your average cost in USD to solve one customer chat session multiplied by the total number of self-served chats by the AI. 

This metric shows the savings that come from using Conversational AI in customer service. Without the AI this amount of cost would have been spent on human labor to answer frequently asked questions. With AI the same customer agents’ time can be spent on solving more challenging and complex problems and conducting sales.


Allows to measure the ROI for the customer service automation platform that provides the Conversational AI solution and the Intelligent Virtual Assistant. Here you can find more information on price ranges of Conversational AI and chatbot solutions.

Average benchmark

Depends on the volume of deflected chats and the average cost of 1 solved customer service interaction. Monthly savings with larger customer volumes can reach $100,000+.

3. Positive Customer Feedback Share


Percentage of the number of positive feedback to conversations with AI divided by the number of all given feedback to the AI. For each conversation customers can leave a positive “thumbs up” feedback or a negative “thumbs down” feedback. 

Based on this a user chat session with the AI gets its feedback mark. After dividing the positives with the number of total feedback given, we get the positive customer feedback share. 


Shows the explicit happiness that customers have with the AI. It is a clearly defined metric and measures directly how customers have rated the customer service AI solution with a simple yet powerful feedback method.

Average benchmark


4. AI Impact On All Customer Service Channels


The number of self-served users by the AI that did not follow-up with a call or did not send an to the customer service call center. 

This is an impactful metric designed to show how the deflected customers did not call into the call center or did not follow-up with an email. Can be used when Conversational AI has authenticated the user and the self-served chat session ID is cross-referenced with inbound authenticated call center calls and emails.


Show the net impact of Conversational AI in messaging across all inbound customer service channels (phone and email).

Average benchmark

Depends on the setup of the customer service automation platform and customer service channels used.                                                                                                                                                                                                                                                                                                                                        

What Is A Chatbot?

A chatbot is an interface through which the user can obtain information automatically from a software program. The UI is usually in written form (chat, messaging) and often the chatbot presents the user information with a simple multiple choice selector. Advanced AI chatbots (also known as Intelligent Virtual Assistants, a subdomain of Conversational AI) are capable of providing intent detection and resolving customer inquiries through natural language understanding. Chatbots are used in multilingual customer support, lead generation and sales automation across a variety of channels (e.g. websites, Facebook Messenger, WhatsApp, SMS).

What Is An Intelligent Virtual Assistant?

An Intelligent Virtual Assistant (or IVA for short) is a machine learning based system capable of understanding human language and answering questions that people ask. These intelligent AI assistants are not only capable of presenting a multiple choice selection of answers to the user but also understand user intent from free text. Some platforms also have no code options to make it easier to build IVAs.

As the IVAs are used extensively in customer support automation they are also known as Virtual Customer Assistants. Understanding natural language is a significant leap in the development of smarter machines. The ability to conduct Natural Language Processing gives machines the capability of understanding users in a better fashion and serve more useful and accurate information.

5 Best Intelligent Virutal Assistant Software Tools

1. AlphaChat


AlphaChat is a chatbot software platform allowing anyone to build Conversational AI bots for their website. Aside from Natural Language Understanding, the AI is capable of authenticating users with deep automations for online customer service.

Top Features:

  • Built-in Natural Language Understanding for training Conversational AI.
  • Dedicated tool for website chat automation. Live chat built-in.
  • Use API to authenticate users and provide personalized automated support.
  • Pre-trained industry templates. Works in all languages.

What’s special about this tool: Powerful tool with no-code interface for anyone to build online customer service conversational AI. Launch a bot in minutes. API available for advanced users with deep insight into bot content and AI suggestions on how to improve the Conversational AI. SLA and Alpha Operating System with custom code available for enterprise chatbots. 

Pricing: 10-day free trial. Paid plans from €399/month.

2. Ultimate.ai

Ultimate.ai is a Conversational AI software for automating customer support.

Top Features:

  • Virtual Assistant training.
  • Enterprise focus.
  • CRM integrations.
  • Custom machine learning models.

What’s special about this tool: Ultimate.ai allows building AI bots in several languages and launching customer service automation within weeks.

Pricing: Paid plans starting from €499/month.

3. Aivo

Aivo is a Conversational AI and chatbots building tool for omnichannel customer service automation. 

Top Features:

  • Live chat and co-pilot mode.
  • WhatsApp integration.
  • Voice integration.
  • Virtual Assistant training.

What’s special about this tool: Aivo provides a Conversational AI training engine and connections to multiple platforms including Genesys, Zendesk, Zapier, Slack and Salesforce.

Pricing: Paid plans starting from $99/month + $0.15/conversation.

4. Meya

Meya is a Conversational AI chatbot program for developing customizable virtual assistants for online customer support automation.

Top Features:

  • Multi-cloud hosting.
  • 3rd party integrations.
  • Own markup language.
  • Customizable widget.

What’s special about this tool: Meya offers its own Bot Flow Markup Language for developers in building their Conversational AI.

Pricing: Paid plans starting from $799/month.

5. Verloop.io

Verloop product allows creating chatbots and Conversational AI for customer service automation and marketing chatbots.

Top Features:

  • E-commerce features.
  • Bot builder. 
  • Analytics.
  • Live chat.

What’s special about this tool: Verloop has focus for e-commerce companies seeking to automate support and increase sales.

Pricing: Pricing available on request.


Measure your Conversational AI. Whether it is self-service rate or the costs saved. Most top Conversational AI vendors and some top chatbot vendors provide for this data. It makes your customer service efficient and your customers happier. And you can always analyze where are the opportunities for improvement for the chat automation software.

If you are looking to try this out for your own business or needs, feel free to sign-up for an account with alphachat.ai In just 5 minutes you can get your own natural language understanding Intelligent Virtual Assistant that you can connect with your website. If you have any questions, just reach out or schedule a free 30-minute Zoom call