Gartner Predicts In Five Years Chatbots Will Be The Primary Customer Service Channel
This will only be the case if chatbots are designed correctly...
According to Gartner, chatbots will be the primary customer service channel within five years…but it needs to be emphasised, only for 25% of organisations.
Customer service and support (CSS) leaders are struggling to identify actionable metrics.
~ Gartner
Measures For Success
I have written about practical measures for success in the past, the best approach is to automate current processes. These current processes can include customer calls to a contact centre or live agent chats. The same metrics used for success in those engagements can be used to measure the success of a voicebot or chatbot.
For instance, when replacing an IVR with a voicebot, the same measures of containment, returning customers, NPS and CSAT should be applicable.
Please follow me on LinkedIn for updates on Conversational AI and more. 🙂
🚀 Strategies For Success
Gartner advises that organisations should focus on:
1️⃣ Chatbot Deployment Strategy
The chatbot deployment strategy should be focussed on using existing customer utterances and conversations. These conversations represent customer intent, hence conversations customers want to have. When these customer conversations are clustered in semantically similar groupings, in essence intents are defined. These clusters need to be tweaked according to granularity and cluster size.
This process ensures that business intents are aligned with user intents. Gartner advises that too many chatbot projects start with technology choices and not analysing existing customer conversations.
2️⃣ Containment
Containment is obviously important, as this illustrates the effectiveness of problem resolution within the chatbot. But containment can be artificially increased by not offering agent escalation to the user. In this instance NPS and CSAT scores keep forced containment in check.
3️⃣ Identify Relevant Chatbot Metric
Chatbot metrics should include:
🎯 Net Promoter Score needs to be high.
🎯 Customer Satisfaction Score needs to be high.
🎯 Containment needs to be high.
🎯 Returning Customers via the same or other mediums need to be low.
Read more about metrics here.
4️⃣ Review Cadence
There should be daily transcript reviews attended by the whole squad, joint transcript reviews build a collective and shared understanding within the squad of how the chatbot is fairing.
Transcript reviews also give insight into the true customer experience, literary word-by-word. Quick, small iterative improvements are key, within sprints ample capacity must be allocated to maintenance and optimisation.
Please follow me on LinkedIn for updates on Conversational AI and more. 🙂
In Conclusion
It cannot be stressed enough, that prior to launching the conversational agent, any existing customer conversations or utterances must be analysed and intents and entities derived from it.
This process of automating existing customer conversations will make it easier to establish realistic and relevant metrics for success.
After launching a chatbot or voicebot, the improvement of the chatbot pivots on an ongoing process of analysing customer conversations to improve the experience.
The hard work only starts after launching the conversational agent…
I’m currently the Chief Evangelist @ HumanFirst. I explore and write about all things at the intersection of AI and language; ranging from LLMs, Chatbots, Voicebots, Development Frameworks, Data-Centric latent spaces and more.
