Would you trust a chatbot to mix you a cocktail?
Imagine you’re standing at a bar and the barman comes over to serve you.
You: Hi, I’d like to know which cocktails are included in the happy hour, please.
Barman: Sorry, I don’t understand.
You: I’d like to know which cocktails are part of the two-for-one deal.
Barman: Would you like a drink?
You: Yes! How about a martini, is that 2-4-1?
Barman: 241 martinis coming right up!
Wow. That’s a huge – and hugely frustrating – round of cocktails you just ordered. You leave before it gets messy – and you vow never to return.
You might well be familiar with this kind of experience – even if you don’t drink cocktails. I know I am! This fairly useless barman is like a poorly built chatbot. The kind that enrages people like me on a variety of websites.
Although they are built with intelligent virtual assistant technology, these chatbots don’t seem to be very intelligent. They can’t understand simple questions. Their tone of voice is awkward or annoying. And they can have a huge impact on customer experience.
Research by Salesforce found that 80% of customers say the experience a company provides is as important as its products and services. So, if that experience is frustrating, customers will go elsewhere.
The benefits of digital assistants
Artificial intelligence (AI) platforms have changed the way that customers interact with brands.
I’ve seen many companies introduce digital assistants to help save money and time when it comes to dealing with customer queries.
A well-built, well-implemented chatbot can bring a range of benefits to both customers and contact centres. Chatbots can:
- answer customer queries 24/7
- understand and respond to questions in multiple languages
- improve CSAT and NPS scores – because they answer questions very quickly and accurately
- reduce enquiries to a contact centre – Sabio customers have seen enquiries reduced by 30%.
Chatbots can also boost morale and retention of contact centre staff. How? By handling the steady stream of basic and repetitive enquiries, such as password re-sets and payment details. Contact centre staff can then get their teeth into more complex queries that require in-depth skills and expertise.
But for a chatbot to work, customers need to enjoy interacting with it – or at the very least, trust it. If you’re going to ask customers to change passwords or hand payment details over to a virtual assistant, it needs to feel safe to do so.
What to consider before building a virtual assistant from scratch
Some businesses turn to their IT team to build a chatbot. In my experience, this is where the issues arise.
Whilst an IT team can provide essential technical expertise, you need a blended team to create a bespoke chatbot that meets your objectives, boosts your brand and delivers on customer experience.
Building and deploying an intelligent virtual assistant is not an IT project – it’s a customer experience project.
Before you build your own bot, you need to consider:
- Do you have UX experts with experience of building conversational assistants?
- Have they built virtual assistants for your sector and target audiences before?
- How does a virtual assistant fit in to your wider customer journey?
- What happens when a virtual assistant can’t answer a customer query?
- How will you ensure it looks and feels like part of your brand?
- Who will optimise and develop the technology long-term?
Watch the video and hear everything you need to consider before building your own Virtual Assistant.
Digital assistants must feel trustworthy
Contact centre staff do not simply process transactions. Seventy-two per cent of service agents say their interactions with customers are relationship oriented (Salesforce, State of service, Third edition). Agents are there to offer support when someone has a question – or if something goes wrong. It’s their job to empathise, reassure, build rapport and, ultimately, find a solution.
So, if you’re hoping to build a brilliant chatbot that supports your contact centre, you have to build one that’s equally reassuring. You need an intelligent virtual assistant that can interpret how your customers are feeling and responds appropriately.
It’s no good feeding a chatbot some standard questions and answers. Because every customer is different, every question will be different. So, every response should be different, too.
Using machine learning to build rapport
Sabio’s virtual assistant uses Natural Language Processes (NLP) to create useful, helpful answers. NLP is based on deep learning that enables computers to gain meaning from user inputs. We humans listen to a question and then give the most appropriate answer, depending on who we are speaking to. We might modify our responses to a child, a colleague and a customer, for example.
In the same way, Sabio’s virtual assistant solution assesses what the meaning of a customer’s question is. It analyses the context and then creates an appropriate response.
The dialogue also feels natural because it is constructed by our experienced team of dialog designers and ux specialists. When you install Sabio’s virtual assistant solution, this expert team can work with you. They’ll ensure that it’s programmed with common words for your sector and has the right tone for your brand. And as this form of Artificial Intelligence (AI) evolves, they’ll continue to optimise the virtual assistant so that it meets your customer’s demands.
When it comes to Virtual Assistants, Sabio can help
If you’re considering a virtual assistant, we can partner with you to create one that reflects your brand, boosts your productivity and helps your customers.
If you’ve already launched a virtual assistant and need some help, we can improve it.
We can work with you to ensure that a virtual assistant is introduced in the right way, on the right channels, to the right customers.
Now, that sounds like something worth celebrating. Pass me one of those martinis – cheers!
Request a Virtual Assistant Demo