When chatbots first arrived on the customer service scene, many thought they could be used to replace human agents. Their ability to talk and learn made us believe that they could perform the role of a virtual agent, resolving more customer queries unassisted.

By Ryan Falkenberg, CEO of Clevva

Well, we all know that has not happened. On the whole, chatbots have been a major letdown. Aside from their inability to understand most requests, their capability to do anything helpful remains frustratingly limited.

Most chatbots are what are known as info bots or Question Answer Machines. They are effectively smart librarians, skilled at predicting an answer to a customer’s question.

Unfortunately, the relevance of their answer is dependent on the quality of the knowledge base they have access to. And in the case of most companies, this knowledge base is not very granular. As a result, most answers tend to be generic, even if your question is very specific, and you end up having to speak to a human agent anyway.

As chatbot technology has improved, many have learned to perform basic requests e.g. top up my airtime by R50. These transaction bots are useful when the requests are simple and clear. They struggle as soon as you are not sure, and need advice. This is the case for most service queries, issues and complaints.

As a result, most service conversations that start with a chatbot end with a human agent. Very few can be automated straight through.

Fortunately this has changed with the introduction of service bots. These digital service experts are capable of rich service engagements that result in known queries, issues and complaints being resolved without having to involve a human agent.

Service bots aren’t simple digital assistants that wait for a clear instruction before trying to perform a required action. They are intelligent digital advisors that help you clarify your request, assess your situation, analyse your needs and diagnose the root causes of your problem before recommending a solution or taking action.

This means customer service teams can now assign different work to different agents. They can allocate significant volumes of tier 1 and 2 queries to their virtual agent team. These queries get resolved via multiple digital channels, such as the website, mobile app, WhatsApp and email channels. The human agent team gets assigned tier 3 queries, as well as any queries that demand greater levels of empathy; where culture and language are important; where humour helps build human connections; and where value is sought from the human experience, not just the resulting answer.

This changes how call centres hire and train human agents as you aren’t looking for tier 1 skills capable of performing basic triage on customer queries. Your virtual agents can handle that. You’ll instead look to hire agents with more in depth skills – both in how they handle customers with different personalities, cultures and languages; and the level of expertise in your particular product or service. It likewise means you can focus on training your existing teams on delivering more sophisticated service to more complicated queries.

Using service bots to perform the role of virtual agents will also reduce your total cost to serve, while improving the quality of the resulting customer experience. Customers will get more done digitally, and will only ask to be connected to a human agent when they need the human touch. These engagements will be more valuable and more impactful. They will also drive up customer loyalty.

Service bots allow customer service teams to realise the vision of a virtual agent capable of resolving queries, just like a human agent. As a result, human agents are increasingly being liberated to build more meaningful customer relationships, and to re-humanise the customer engagements that really matter.