Artificial intelligence is a hot issue for customer service teams. The idea that AI-powered bots can cover the majority of customer interactions is an attractive one, in terms of efficiencies. However, chatbots need to be used carefully. Similarly, there are multiple places where AI can be used within your wider customer engagement strategy.
According to research by the Institute of Customer Service, there’s a good degree of understanding around AI already – around 65 per cent of customers could name AI applications when asked, and more than a quarter had already used common AI assistants. For most of those surveyed, chatbots seem to be a valid route to getting information or having their queries dealt with.
However, it’s not as simple as simply setting up a chatbot as a new channel. For companies looking at how to use AI effectively, there are some quandaries to consider. How can you use AI-driven services as part of your wider approach to customer support? Can you consolidate your management of customer interaction across multiple channels? And how can you design your processes to make the interaction as frictionless as possible, whether a chatbot is involved or not?
Planning ahead around services and bots
Based on these survey results, there are several steps that you can take around chatbots and service:
- Understand your customers and their preferences for channels to use
Your customer base may be happy to use bot services rather than phone, online chat and email, as they find chatbots easier and faster to deliver answers; alternatively, they may see chat as being too impersonal. Looking at demographic data can help here – younger customers tend to be open to using technology more than direct personal contact – but it is best not to assume this as fact.
Running a focus group or customer survey on preferences for channels to use can help if you can get a representative sample of responses. This kind of activity can show that you are listening to your customers, and looking to meet their needs in the most appropriate ways.
- Start small with chatbots
During the launch of your chatbot, you may be tempted to start all your interactions with customers using this new service. However, you should start this chatbot service as an option to start a conversation For those that are happy to use this service, they can opt in to a conversation; equally, the bot can automatically route other customers through to another channel.
- Support chatbots with the right content strategy
Chatbot services rely on providing access to the right content for customers to solve their own problems – for example, sharing FAQs and knowledgebase articles, connecting customers to website pages, and answering those queries with specific data based on common searches. However, those criteria will change over time. Supplying new articles – or writing new ones that answer new questions – will be an important task for customer support agents in the future. Without this content, your chatbot will lose its value over time.
- Integrate your chatbot into your wider help desk system
This should be obvious, but your chatbot will be less valuable if it runs independently and cannot provide information through to your agents. If a customer has a difficult problem – or one that requires more nuanced understanding and empathy – then a chatbot will probably not be able to close the issue alone. This will mean a hand-off to a human agent, either within the same chat window or on a call.
When this takes place, the agent should have the whole chat record and background open to them in one go. If not, they run the risk of asking the same questions and making the customer feel frustrated. Chatbot programmes should be able to automatically populate and manage tickets in your help desk system, rather than having staff spend time on setting new tickets up themselves.
This is an important requirement, because it will demonstrate that your approach is working. Without the right data on chatbot activity in your help desk system, you won’t be able to get an “apples to apples” comparison on how well you are doing, or how many tickets the chatbot is deflecting to a self-service channel.
- Plan to use AI to power other services too
AI is not just about chatbot support, although the two technologies are intertwined. AI can also be used within other service channels too – from recommending content to human operators based on context of conversations, through to more advanced process automation that can take time out of interactions.
Chatbots will probably be the most obvious use of AI to start with, but don’t look at AI as solely about chat. Instead, look at how you can gain the most from AI across all your customer service channels over time. By keeping this approach in mind from the start, you can avoid being locked into different services for different channels.
AI has massive potential. Chatbots can help you serve more customers and answer simple questions quickly. Over time, AI can take on more complex questions and meet customer needs in a way that fits with how many customers want to be served. However, without the complete picture over all your service channels, you run the risk of creating new silos for customer support that only meets some of your needs, some of the time. It’s important that AI is used as part of your overall customer engagement strategy and with a defined goal in mind.
My role at Freshworks as General Manager UKI is to manage the company's operations, including strategy building, developing our own high impact sales and support teams, and managing our performance.
Freshworks provides IT Service Management, Help Desk and Customer Relationship Management products. Currently, Freshworks helps over 150,000 companies worldwide deliver customer support and IT Service Management via email, phone, chat, web, social media and mobile apps.
Outside work, I take great pride in my kids and support Everton FC.