What is the importance of customer experience (CX) and virtual assistant implementation? This blog interviews Matthew Doel, Managing Director and founder of EBI.AI, to discuss the differences between chatbots and AI assistants, AI assistant implementation and its importance for providing truly seamless and engaging customer experience.
This article covers:
- Why is positive customer experience so important?
- What are the advantages of a fully managed AI solution over build-it-yourself type bots?
- What are the common challenges and considerations companies have in maximising the value of their AI assistant?
- For a company who has implemented or is considering implementing AI, what is the expected overall impact on the business over the next 10 years?
- How can companies incentivise customers to use their AI assistants?
One quick Google reveals thousands of links describing how to build a bot in 10 minutes. It’s perfectly fine if you want to play around with AI technology and are genuinely just curious how far you can go creating your own technology assistant. However, if legitimately implementing a bot within your customer experience (CX) journey, it’s time to get serious about the CX problem you are trying to solve.
Your CX problem completely dictates the design of your virtual assistant and if it fails to engage your customer – you risk losing that customer. Although seemingly common sense, there is a lot of hype around chatbots and AI implementation, yet less clear differentiation between the different types of virtual assistant implementation that impact how you solve that CX problem. We seek to distinguish between simple rules-based chatbots, build-it-yourself AI bots and fully managed AI assistants.
When was the last time that you set a kitchen timer without Alexa or Google Home’s assistance? In reality, voice AI assistants are trained to do a whole host of tasks, yet they have become the norm within households due to their prowess in performing a few useful tasks, really, really well. Online however, many consumers still hold a negative view of AI assistants due to previous bad experiences with rules-based chatbots and the fear of the unknown.
Allow me to be crystal clear, an AI assistant is worlds apart in technological capability from the rules-based chatbots of the noughties that captured customers in a never-ending and frustrating loop. Rather than writing limited and ineffective rules in response to common consumer queries, the AI chatbots of today use Natural Language Processing in relation to specific customer use cases.
For AI assistants to provide true value to customers, the first step is the most important: understanding where in the CX journey a virtual assistant can transform customer experience and increase engagement. Only by solving real customer use cases quickly and effectively can customers begin to trust and incorporate AI chatbots within their interactions with your company and forget the bad bots of the past.
Why is positive customer experience so important?
Customer experience is the impression your customers have of your brand as a whole throughout all aspects of the buyer’s journey. It results in their view of your brand and impacts factors related to your bottom line, including long term revenue (Hubspot, 2020)
In fact, companies that prioritised and effectively managed customer experience were 3X more likely than their peers to have significantly exceeded their top business goals in 2019 (Adobe, 2020). This commitment to customer experience is where badly thought out, built-in-10 minutes-bots are doing customer experience a dis-service. We turn to Matthew Doel, Managing Director and Founder of EBI.AI to tell us why.
In your experience, what are the advantages of a fully managed AI solution over build-it-yourself type bots?
“When we started out, we probably took the same approach as the build your own bot process. There’s a series of steps: put simply, you connect a natural language service and a user interface, write a bit of code and pull it all together to build a bot, but it’s a basic technical solution at that point. Often what you’ve hacked together is a bot that won’t meet regulatory security and audit requirements, is difficult to maintain and not scalable.
There’s a huge engineering task between that 10-minute build-a-bot to a well-organised platform that can cope with launching and running virtual assistants at scale. And the conversation itself, the business use, in terms of what the customer experience is going to be, has been given zero thought at that point. At EBI.AI we have spent 6 years taking that initial bot build through to an enterprise class platform that, from an engineering perspective, has all the things in place that you’d expect from a high-quality SaaS solution. The typical FAQ style bot is very much the Minimum Viable Product (MVP) of what it takes to get a project up and running, we can now take that bot to the next level to actually offer something beneficial to the customer.
Now that we have done all the hard work on the engineering side, we can now prioritise what the customer experience is going to be interacting with that bot. We work towards what are our customers looking to achieve and what is the nicest possible experience we can provide to their customers. Our team who are writing the conversations for our virtual assistants have got Psychology and Linguistics backgrounds and they write in a way that the conversation flow is both enjoyable and as effective as possible. Rather than ask lots of follow up questions, they will try find the path to the answer in the least number of interactions so that the bot is both quick and efficient.
Additionally, when it comes to user interface, rather than give 10 text-based options, we might show a visual carousel that gives the customers different options to click through. These widgets allow for a richer and engaging customer experience because after all, a picture can still paint 1000 words.”
What are the common challenges and considerations companies have in maximising the value of their AI chatbot?
“I think the key thing is to really look again what they are looking to achieve from implementing an AI solution. The common problem to be solved, particularly during Covid-19, is a customer service use case where a customer has an issue or a question and the current journey includes a poor digital experience that eventually defaults to a long wait time on the phone, to get in touch with an under-staffed and overwhelmed contact centre.
However, we are increasingly dealing with forward looking companies who are looking to increase customer engagement and life-time value (LTV). For example, for our client Coop Sweden, one of Sweden’s largest supermarkets, we are building out customer engagement opportunities where it might be finding a recipe with what’s left in a customer’s fridge, or asking how busy it is in their local Coop. By building this into an AI assistant called Cooper who lives on their website and app (and soon smart devices), we’ve given Coop the tools to be a helpful concierge assistant throughout their customers’ day to day lives. These engaging customer experiences are sticky and keep customers coming back, directly impacting LTV.”
For a company who has implemented or is considering implementing AI, what is the overall impact on the business over the next 10 years?
“Unlike traditional computer systems which are at their best on the day that they are implemented, the opposite is true for AI. From day one, the AI is learning and continuing to learn. If the correct business systems are in place to capture those data insights and knowledge, then the company will continually improve the customer experience they give.
In terms of the overall business impact, what we commonly see is a move towards an AI data-driven business. Once they’ve gotten their heads around the capabilities of this technology, an AI solution typically becomes a bigger project that can transform the way the business uses data.
A basic example, going back to our customer Coop, the machine learning algorithm within the customer data platform will be able to identify how frequently a customer buys for example, ketchup, which often varies from customer to customer. By using the data trends that the machine learning algorithm identifies, Cooper now has the capabilities to suggest customers buy a new ketchup just as their previous bottle runs out. This is a prime example of how a company can maximise the value of their data by using an AI solution to personalise experiences that increase engagement, basket size, frequency of order and ultimately life time value.”
How can companies incentivise customers to use their AI assistants?
“I think there’s two things here, as customers have more and more great experiences with AI assistants, they will use them more. However, for the companies it is about taking risks and giving incentives for customers to use their AI assistants themselves. For example if a customer calls and it says the current wait time is 30 minutes, state that you have a new automated service that they can try out and if at any time they feel uncomfortable, they can press 0 to return to the 30 minute queue. Word it in such a way that there is no harm in giving the AI assistant a go and bet on the fact that they will have a better customer experience.
Similarly try to befriend your customers. If your AI assistant hasn’t been able to answer them, explain that you’re very sorry that you couldn’t help, that this has been logged and in the future (so long as their query was in the realm of the customer use case) the AI assistant will be able to answer this query because it will be added to its repertoire.
Finally, we are finding adding conversational interfaces to existing user experiences to be particularly potent. We have added a conversational interface front and centre for two of our clients within their apps now. This allows customers to try the conversational AI assistant blended with the traditional app experience that they are used to so that they can find the right balance that works for them.”
- Conversational AI platforms are the next level up from build-it-yourself bots #banishbadbots
- Understanding your customer use cases for AI assistants is crucial to maximise their value
- AI is more than automation; it’s engaging and personalised customer experience
- Incentivise users to use your AI assistants and #banishbadbots
Thanks for reading this interview with the Managing Director of EBI.AI, Matthew Doel. If you want to know more about our AI platform and what we do, please request a demo and we’d love to chat with you about how to banish bad bots and transform your CX with a Lobster AI assistant.