Creating chatbot KPIs (key performance indicators) is the difference between innovating for the sake of it or being invested in what chatbots can offer to your business and brand. Establishing the right KPIs for your chatbot is the way you provide a customer experience that blows your competition out of the water.

But what are the best KPIs for your chatbot – or, for that matter, your digital human? We asked those who have deployed chatbots and are well on their way with creating a digital workforce.

Here’s a rundown of what you’ll need to know.

How to create KPIs for your chatbot

You’ll first want to determine what “success” will be once you put your bot out into the digital wild. You’ll then set some ways that you’ll measure how close you’re coming to that success.

Will you consider the sheer number of interactions your chatbot has as success, as you aim to build awareness of your chatbot channel? How will you measure and report on that in the backend of your chatbot platform?

Or is success at this stage tied to the bottom line? In which case you’ll want to measure ROI, or the cost savings made by resolving issues on the chatbot level instead of it being escalated to a human support level.

It’s important to keep in mind that your chatbot KPIs and digital human KPIs might differ. Just as your different sales, marketing and support channels have their own determination of success and ways to measure it, so too do your chatbot and digital human workforces. 

We’ll touch upon some of these differences shortly.

Which chatbot metrics matter most?

It probably goes without saying that the number of KPIs businesses use to track success differs broadly, as do how specific these metrics can be to your goals. 

So when when we asked chatbot strategy leaders, employers and enthusiasts what single metric they consider most important when measuring the success of their chatbot, there are a lot of possible answers. However, four categories came through:

  • The sheer number of interactions the chatbot has
  • The speed at which a resolution is reached
  • How satisfied customers are with the interaction 
  • And how the chatbot reduces the cost of serving these customers. 

What these brands find most important is clear. Our polling shows that customer satisfaction takes an easy lead with 48% of the vote. Reducing cost to serve those customers comes second (19%). Which is closely followed by speed of resolution and the number of interactions the chatbot makes (both tied at 17%).

Chatbot KPIs vs digital human KPIs

Drift and Heinz Marketing found that chatbot usage almost doubled between 2019 and 2020 – growing from 13% of brands using them to 25%. 

And yet 82% of consumers say they want more human interaction with brands, not less. That’s not to say your chatbot strategy is wrong; it just needs more humanity.

If you’re new to our blog… hi. We’re a digital human company whose platform humanizes chatbots and lets your customers have direct “face to face” communication with their human form – a digital human.

Not only are consumers after a more human experience, but brands are too. In fact, 42% of people we polled said their number-one priority for their strategy was creating a more human chatbot experience.

With a digital human, your chatbot drives the conversation, but your customers get the satisfaction of being seen, heard and valued through the power of human interaction.

So, when it comes to setting KPIs for your chatbots and your digital human, things can change slightly, because of the nature of each channel. 

It’s possible for chatbots and digital humans to have some similar KPIs. For instance, boosting customer satisfaction, reducing cost to serve and improving the speed of service are all benefits of both forms of automated conversational AI.

But it’s important to not forget the fun – the part of customer experience that extends beyond mere satisfaction.

Noel Leeming (part of the largest retail group in Australasia) employs Nola, who plays two roles. Nola is a UneeQ digital human who acts as a concierge in Noel Leeming’s flagship store, and she also takes the form of a chatbot on the brand’s website.

While both channels measure the number of interactions and the quality of those interactions (the number of handled vs unhandled requests), Nola as a digital human has another job – adding a greater experience.

As Dylan Weymouth, Business Operations Lead at Noel Leeming, explains in our eBook: “our Nola chatbot is a quicker form of service, and is direct and to the point. 

“As a digital human, Nola is there for engagement – for users to chat with. If she can solve a problem, that comes secondary to us. We want people leaving engaged and thinking “that was fun”. You can have more off-topic and random conversations with her, and she’s designed to answer more “human” questions.

“We look at ‘fun’ measures of success, like how many people take a selfie with her.”

Going beyond chatbots and towards the digital workforce

As you think about your chatbot KPIs, consider which measures will determine the quality of the customer experience.

Are customers using a chatbot because they don’t want to wait to speak to a human, or taking up valuable human resources time because they don’t like speaking to a chatbot?

If there’s nothing in between, you’ll perhaps struggle to set a viable customer experience KPI for your chatbot. For that, consider what a digital human can add to your digital workforce.

To find out much more, including a deep dive into the actual consumer pain points solved by a digital workforce, simply download our free eBook below.