By James Penfold, Head of Strategy & Experience Design at creative technology agency Rehab.

There’s a chatbot for everything, so it seems. According to Acquire, 1.4 billion people are willing to use chatbots, and 74% of users prefer to talk to chatbots while looking for answers to simple questions, says PSFK.

The positives of chatbots are that they’re (usually) easy to navigate, cut time and resources, and provide quick fire solutions for customers who need an answer there and then. In fact, they can answer up to 80% of standard questions (IBM). However, there is a huge opportunity for more success for businesses in the form of customer relationships, upselling and customer feedback – which can help shape future business and product decisions, but there are a lot of improvements to chatbots that need to be made in order to advance this model.

A big advantage of chatbots is that they can help alleviate pressure on staff, triaging customers’ questions and solving problems immediately without the intervention of a human. However, often chatbots can provide more confusion and frustrations, as these models are not always designed to the best of their ability. Companies are using these systems as a way of ticking boxes, rather than providing a simple, real time solution for their customers, which ultimately leads them to wanting to speak to a human to solve their problem.

The user experience of chatbots can also be deemed as off putting and sometimes irritating. Often when browsing on websites, as soon as you land on the home page, a chatbot immediately pops up asking if you need any help. These pop ups can slow down the page, take time away from browsing the site and could encourage users to leave the site altogether.

Customers want a smooth browsing experience, without the interruption of pop ups, meaning users can be susceptible to banner blindness, where subconsciously they will ignore any banner-like information that appears on their screen. This doesn’t mean chatbots cannot be useful to their customers browsing on their site, however companies need to consider the UX of their website when integrating chatbots, to ensure they are a help not a hindrance.

In order to see real improvements to the chatbot experience, companies need to capatalise on data capture to create an overall personalised experience. Conversations with chatbots need to be contextually relevant, and for this to happen they need to know what the customer is really looking for, and not just provide general information that is suitable for anyone who accesses a site. Where has the customer come from? What products have they been looking at? Have they recently placed an order? This kind of information is extremely useful in creating an optimal user experience, and enables the chatbot to provide answers quickly and effectively.

As technology rapidly advances, customers are beginning to expect a similar in store experience, online. As if a store assistant knows it’s regular customers, digital footprints and data capture can allow chatbots to know almost everything about customers’ online shopping habits, and design their interaction with them based on this information. Machine learning can also step in here, and add an element of sentiment analysis to understand the users feelings towards their query. If they are angry about their problem, the chatbot can detect this and interact with the customer in a different tone that provides a VIP experience and will lead to the customer using the chatbot again in the future.

Chatbots can be used across almost every industry, from financial services such as Snoop, the digital advisor that helps you make smarter decisions with your money, to beauty brands and everything in between. At Rehab, we recently created a chatbot for high end beauty company, Estee Lauder, called Liv – which took well known products and educated users on how to use them and advised on customers’ beauty regime as a whole. With Liv, we were able to explore how Whatsapp can be used as a new CRM channel to provide opt-in first party data. Liv provided a 7-day course of skincare tips and tricks, delivered directly into the conversation interface and built with customisable reminders, which allowed users to build and maintain an evening skincare routine that suits them. From product advice audio, tutorial videos, links to further support and discount vouchers, to a conversational tone of voice packed with emojis, the rich variety of content contains the perfect recipe for an authentic experience. It became part of users’ routines creating a more human and personalised experience that customers could rely on, which I predict more and more chatbots will be doing in the future. The value of Liv was highlighted with the level of repeat engagement. During the pilot, over 20% of users who started the experience on Whatsapp continued the conversation with the brand daily for 7 days.

Over the next five years, we expect to see big advancements in the chatbot world. Menu type responses will be a thing of the past, with customers feeling like they are interacting with a human through advancements of AI and machine learning. The level of accuracy will also increase, with chatbots knowing exactly how to respond to challenging situations only humans can currently tackle. Another element we should be seeing more of, is voice communication with chatbots. Customers will be able to speak to their chatbot, make easy requests and even purchase items using their voice. There are already examples of this out there, for example Amazon Alexa. However, as consumers build their trust with chatbots, they are going to be more likely to make small, convenient purchases, reducing the need for call centers and telephone operators.

To conclude, we have come a long way with chatbots in the past 10 years, but there is still a long way to go.

To be successful, gain consumers trust and free up valuable resources, chatbot experiences need to mimic human interactions, and produce solutions similar to what real life people would provide. In order to achieve this, data capture is the key. Companies need to know their customers, online habits and issues to provide comprehensive and personalised solutions to their problems.

Chatbots certainly won’t be replacing the human workforce, but a hybrid approach will allow for a more efficient way of working, as long as the technology is utilised correctly.