20.01.22 | 7 mins

Talk to your machine

Conversational design tendencies in the age of relationships.
By Vicente Ocaña, Clara Sánchez-Puga, Nacho Rodríguez


It is interesting how quickly we get used to speaking to Alexa or Google’s assistant. Far from the uncanny valley of technology that imitates humans, we find it extraordinary to give voice orders, or even better; get answers and initiate a conversation.

In order to achieve conversations that feel as human as possible, they are created (and iterated) from conversational design, a discipline that merges design and voice interface, interaction design and UX writing, among other possibilities.

The core of it all is natural language processing (NLP). This is one of the components for artificial intelligence: it is a computer’s ability to understand oral and written human language. NLP was born about 50 years ago and stems from the field of linguistics. It also contains numerous day-to-day applications beyond asking Alexa about the weather; it ranges from medical research to business intelligence, where new opportunities arise thanks to the new journeys that are based on data, different contexts and specific moments.

But what are the differences between a chatbot and an intelligent assistant? They share a common base: an NLP and the capability to interact with the user outside the “point & click” interface.

zapiens_habla-con-tu-maquina_home-pod

HomePod mini, Virtual Assistant Siri


However, intelligent (or virtual) assistants are ahead of chatbots in sophistication and features in some instances:

  • AIs include an NLP with the ability to accumulate knowledge based on the user’s interaction record, and a focus on continuously improving the service. Chatbots, on the other hand, are based on rigid data structures and specialise in responding to a limited set of questions and phrases without a proper analysis, classification or answer prediction.
  • AIs are capable of maintaining consistency, constancy and interactivity across all customer touch points with the service, regardless of the channel, whereas chatbots cannot maintain consistency due to their intrinsic limitation to the chat channel.
  • AIs use ANN (Artificial Neural Networks) to recognise, analyse, classify and predict responses.
  • In this degree, they’re available for the users at all times, learn from each conversation and proactively provide answers. A chatbot is limited to a conversational sphere of information-processing without a progressive learning system, and lacks the features a more complete virtual assistant can offer.


We all know the most common virtual assistants: Alexa (Amazon), Google Assistant, Microsoft Cortana, Siri (Apple), Bixby (Samsung) or Aura (Telefónica). Many companies are incorporating chatbots into their customer experience: Poncho, National Geographic Einstein Bot, Lidl’s Winebot, Kian (Kia Motors), Whole Foods Bot, Duolingo Bot.

zapiens_habla-con-tu-maquina_nest-mini

Google’s Nest Mini


How will conversational design develop further?

We look around to analyse trends:


Personalising the experience

The incorporation of chatbots or virtual assistants in the customer journey is becoming more and more common, not to mention the rise of home voice assistants during lockdown and the toughest months of the COVID-19 pandemic. Conversational design permits a much more advanced customisation of any service, beyond socio-demographics, and leads to new business possibilities.

In this sense, Hey Disney, Disney’s skill for Alexa that has been running for just over half a year now, allows a great experience repertoire, which includes character voices from Marvel, Pixar, Star Wars and even Disney itself. Far from ending there, Hey Disney will be available this year in the Walt Disney World Resort hotels through an Echo Show 5 fully tuned up as Mickey Mouse, with which customers can answer questions related to the day-to-day life of the hotel or theme park.

zapiens_habla-con-tu-maquina_hey-disney

Hey Disney


Learning

A company’s most important asset is its relationship with their customers. An AI system can help humanise the relationship, learning users’ preferences and habits, turning a simple transaction into a positive vital experience through fluid communication, as well as identifying needs and being able to offer new products and services in a personalised way.

The connection between user and assistant/chatbot starts by assessing the customer’s expectations and learning their habits, behaviours and context through data processing. We observe a flourishing trend in learning engines VS engagement so that digital assistants continue to grow and acquire new capabilities continuously and progressively.

zapiens_habla-con-tu-maquina_stable-difussion

User-generated image using Stable Diffusion


Phygital – Metaverse

Recent launches such as Meta point to a trend we have been noticing: beyond human-like virtual characters or avatars, which we expect to see an increase in the coming months or years, we see that customers no longer differentiate between channels or devices. The dichotomy between the physical and digital world has disappeared thanks to mobile devices and rather than choosing between one place or the other, what we observe is a seamless experience that unifies both worlds and enables experiences that coexist simultaneously.

Thus, we will see the growth of consistent experiences that unify all channels and maximise touch points throughout the customer journey, while maintaining coherence and fluidity in the transition from digital devices to physical channels and vice versa. Immediacy, immersion and interaction will be sought.

zapiens_habla-con-tu-maquina_planetas-alex-castro

Illustration by Alex Castro / The Verge

zapiens_habla-con-tu-maquina_meta

Meta Launch Presentation


Anthropomorphising

There are currently a multitude of projects working in the field of emotional and synthetic speech that seem human, and research into how digital assistants manage emotions will be one of the growing trends in 2022.

As aforementioned, emotional speech is more than a virtual avatar and responds to something very clear: the humanisation of the brand is already a necessity. Aura, Sam, Alexa aren’t merely names; they respond to a strategy of a more human experience. We see this clearly in the proliferation of new devices and interfaces with more “human” interaction rules that induce users to attribute anthropomorphic characteristics to the interface: addressing the application as “she” or “he”, saying thank you after the end of the interaction…

zapiens_habla-con-tu-maquina_sam

Sam, Samsung's virtual assistant by Lightfarm Studios


Inclusive design available for assistants

From the trend towards inclusivity in all design, through a population that is increasingly ageing and suffering the effects of the technological gap, we perceive a trend that is still timid but necessary for us to think about how to improve digital experiences for all underprivileged groups, which reach the not inconsiderable figure of 14%.

Proposing an experience adapted to everyone, favouring spoken conversation with patterns of commands that are easy for the elderly, people with speech difficulties, young children, people with reduced capacities or people with temporary or permanent disabilities to take on board and internalise. Designing inclusively is, without a doubt, designing for all.

zapiens_habla-con-tu-maquina_nest-hub-family
zapiens_habla-con-tu-maquina_nest-hub

Google’s Nest Hub


Invisible brand and audio branding

In an invisible brand environment, defining the assistant’s personality perfectly acquires a special relevance. The challenge lies in identifying the brand and customising it through the assistant, which will keep the brand promise in the conversational interfaces that lack the traditional visual elements that identify a company or service.

In this sense, audio brands are gaining importance as one of the fundamental supports for the assistant’s personality and the communication tone used in each moment. Sounds, jingles, tags; these are the new logos or brand colours.

zapiens_habla-con-tu-maquina_her

Scene from the film Her (2013)


Anticipation as a strategy

Working on anticipation is one of the efforts that companies must make in order to achieve a stable relationship with consumers in the medium to long term. We refer to an assistant that takes initiatives to start conversations without the need to be summoned by the user, since to be really useful, the assistant should know and help without receiving any kind of order.

zapiens_habla-con-tu-maquina_blade-runner

Scene from the film Blade Runner 2049 (2017)

Next

Trainers as a cultural icon

What is it about sneakers that we all love?