Booking airline tickets, getting a mortgage, designing trainers or reading favourite news, chatbots are being used by customers for all sorts of things. But how might your organisation use one and where should you start?
(Need to go back a step? Take a look at our beginner's guide to chatbots)
There are three main ways chatbots can help organisations: improving or saving money on customer service; increasing sales; getting content in front of users.
Customer service
Transport for London, Pizza Express, American Express all use chatbots to handle customer service enquiries. Voice assistants such as Alexa or Google Home allow customers to speak to your agent as they would to a call centre. Chatbots on websites & social media allow customers to talk, using Siri or Cortana, or type messages to your agent. So why are chatbots good for customer service?
- Artificial intelligence changes everything. Gartner predict that more than 85% of customer interactions will be managed without a human by 2020. Organisations with large amounts of call, email and social media contacts can have these analysed by machine learning to understand the questions customers ask and the most appropriate response. They can then respond to future queries automatically. The machine learning means the chatbot can respond to questions which have never been asked. As typically over 90% of customer contacts are of similar types these can all be handled by machine learning chatbots.
- Customers want to message companies on social media. A Forbes study in 2017 showed that 65% of customers prefer using messaging apps to telephoning an organisation. Chatbots can be integrated with social media platforms providing an instant response to customer questions in a consistent manner. Calls the chatbot cannot handle can still be referred to a human before a customer gets frustrated.
- Customers expect you to be open 24/7. 1 in 2 customers expect companies to respond 24/7. (Venture Beat, 2016). A chat interface is always open.
eCommerce
KLM flights, Nuvo (mortgages) and Lego all sell online using chatbots. Voice assistants (Alexa or Google Home) also allow customers to buy products in the way they would ask someone in a shop. So why are chatbots good at ecommerce?
- It allows users to buy things in ways that a site’s navigation can’t handle e.g. “I want a good red wine to go with beef”
- It’s where your customers are. Commerce using a chat interface provides unique access to your customers. Customers can also see and buy your products without having to leave social media or their home.
- Users don’t download apps. On average, users engage with only three apps for 80% of their usage (EY 2017). A social media app is typically one of these three apps. Chat interfaces, therefore, provide a gateway to where customers are spending most of their time.
- They’re doing it already. Amazon recently announced that millions of Prime members were voice shopping for gifts, electronic devices and everyday household essentials.
Content
Chatbots provide ways of presenting personalised information to users in ways they want to receive it. The Guardian provides a customised chatbot news service, Cleo provides details of your personal spending and London City Airport provides details of your departing flight. So why are chatbots good at delivering content?
- It’s where your customers are. Facebook messenger has over 1.2 billion regular users. With the continued growth in social media, Chatbots provide an efficient way to engage this audience. Also, voice assistants provide access to your customers inside their home.
- Users are more likely to view your content. A study by Ubisend marketing agency showed that 90% of users viewed business content which was sent through social media. In contrast, industry email marketing open rates are around 25% (Mailchimp, 2017).
- They are more likely to take action. The same Ubisend study showed that 40% of users clicked through to a website after receiving content through social media. By contrast, industry email marketing click through rates are around 3% (Mailchimp, 2017).
Where should you start?
While there are some good examples of chatbots there have been lots of unsuccessful attempts – Microsoft’s Tay which put out inflammatory and offensive tweets or Facebook M which tried to do everything and failed. An early failure puts people off.
But as chatbots could have a massive impact on your revenues, costs and digital presence it makes sense to dip a toe in the water and start experimenting with the technology.
We recommend selecting a business area which will be relative simple to turn onto a chatbot and is fairly self-contained – don’t go for an all singing and dancing application initially.
But choose an area where reasonable financial impact might be gained, whether it is cost saving or sales, as this will demonstrate the power of the approach to the organisation.
But then the key to success is to really understand user’s needs (do some research) then, having built a prototype, test it with users to make sure it works!