Imagine you are visiting an unfamiliar restaurant. The waiter seats you and follows immediately with a “What can I bring you today?”. Do you feel the pressure to respond and stress of spontaneous decision-making?
The same scenario happens with conversational experiences when the bot says “Hi, I am PizzaBot. What can I help you with today?”. Thankfully there is no social obligation to respond in any quick length of time for conversational experiences, however the framing of the question and the multitude of possibilities create a conversational hurdle on the path to decision-making for the user.
Decisions are tough and taxing. We make thousands of decisions every day, from food choices to where we plant our food as we walk. So why do we design conversational experiences with hard decisions?
Granted, every step with a bot will still require a decision from the user, but we can make a users life easier by providing possible paths, wordings, options and even inspire.
To give educated and qualified options you need to anticipate the users desires and goals. Following the previously mentioned example, PizzaBot could follow his introduction with “Do you wish to order a pizza?”, switching from an open-ended to a close-ended question, thereby limiting the number of possible answers. This not only provides for an easier and speedier decision-making process from the user, but also a more tangible amount of training phrases for your conversational AI, yielding a higher likelihood of intent matching.
During our exhaustive analysis of user behaviour with our chat-based conversational interfaces, we have found that providing quick replies (or interactive buttons) is also a very potent way of making sure that the user stays on the happy path.
With relevance being a topic that every product and brand should continuously revisit to begin with, conversational experiences can flourish or falter with focus or neglect of this important keyword, too.
Conversational interfaces have a very special niche in a brands ecosystem of possible touch points and channels and are not always the right answer. This is why it is important to frequently ask yourself the dreaded question: Is a conversational interface the best way for our users to fulfil their desires or reach their goals?
At launch, marketing efforts and novelty will drive users towards you. Curiousity is a critical factor to interface with a vast, yet temporary number of users. To have them return, again and again, you need to create and retain a bond with your user through reliable and relevant services and information. Just like a conversation between humans, conversational interfaces are highly transactional.
Trust is everything in the conversational world. If you don’t trust your conversational interface, chances are you will not return to it in quite some time. That is to a channel that you want available to your users at any point in time, reducing its effectiveness with every user not coming back.
As fragile as trust — a core aspect in human relationships–already is, the damage a simple frustration for a human in contact with a conversational interface can do, is in many circumstances irreparable. Our experience from monitoring over 150,000 messages of humans with conversational AI shows that nowhere breaks a users trust as quickly and decisively as with a bot.
So make sure you keep track of the promises made. Do not over-promise and under-sell. Give the right, qualified answer or be forward and honest about your conversational interfaces skills and knowledge.
In our experience it is simple, yet complex to design and maintain a consistent conversational experience. It is easy to underestimate and compare to other, more conventional touch points but hard to get right, because so much depends on the user.
Users expect to be anticipated. Users expect to be understood. Users expect perfection. Nothing less.
In the case of Lui, our latest conversational interface, the emotional and user-driven chatbot paid full tribute to the primary campaign goal, the acquisition of new customers. Overall, almost 40% of the users who entered Lui’s landing page started a conversation with the likeable, characterful accounting expert. Of the users who talked to Lui, almost 12% bought the product and thus became new lexoffice customers.
With a conversion rate of 4.5%, Lui was clearly above the average of the conventional lexoffice website and 41% above the Black-Friday campaign in the previous year. On Cyber Monday, more licenses were sold than ever before in one day.
Learn more about our conversational design process at brandedconversations.com.