Bringing Aimee into the world: Conversation Design at BT
Specialist Product Designers Qasim Siddique and Francesca Brooking, and Content Designers Jennifer Holcroft and Lucy Storm share how EE’s AI-powered service chatbot came to be.
Conversation Design brings us a unique and rewarding opportunity to create helpful and personalised customer experiences.
Venturing into the world of Conversation Design means bringing together technology, psychology and language to craft natural, human interactions.
It means users must be at the centre of all our work; getting them the answers they need as efficiently as possible, in a way that makes sense to them.
We need to interpret a customer’s query and send back rapid, succinct and useful responses that also feel personable.
To do this at BT, Content Designers with conversation design skills and Product Designers collaborate every day. We bring together UX writing and design, visual design, research, conversational analysis and importantly, an understanding of technical limitations, to create a seamless end-to-end experience to focus on a variety of user needs.
It’s this approach that enables us, as a design team, to bring our digital assistant Aimee to life.
Meet Aimee
Aimee is our AI-powered service chatbot which launched in November 2021.
The decision to create an EE digital assistant was made in response to a growing customer desire to resolve their queries online, and a business need to better serve customers whilst driving down some unnecessary costs.
There was also high demand on our call centres, and therefore long wait times for our callers. An AI chatbot offered us a potential to scale up our digital channel offering and support our customers.
Since launch, Aimee has handled over 3 million conversations. It gives users a self-serve solution, or efficiently directs them to a customer service agent after triaging their issue.
Designing Aimee
When starting our design process, we review high level data including volumes and containment (where Aimee has answered the customer’s question without having to pass to an agent). This helps us identify which customer intent to focus on. Next, we review the current journey if we have one, look at customer conversations, and complete competitor research. At this point, it’s also a great opportunity to talk to operational colleagues to gain their invaluable insight.
An essential aspect of our user-centred design practice is user research. We believe in the importance of speaking to customers early in our process, using generative research to identify the needs and goals of our users. This helps us form a solid foundation to underpin our design work, as well as validating that we’re heading in the right direction.
Of course, this won’t be the only time we talk to customers. We take an iterative approach, and often test small pieces of content and design with users whilst we go.
We’re keen advocates for a build, measure, learn approach (BML). To support this, we recently introduced a Conversation Design Canvas which is used when designing for new and existing intents. It’s based on a resource provided by the Conversation Design Institute, which we have adapted for our needs.
Similar to the BML Canvas, our tool promotes user-centred design thinking at all stages. The canvas encourages us to consider the wider context, setting the scene before we begin designing. We use it to identify customer types, their needs and mindset when entering our journey. The canvas also documents Aimee’s abilities and our business goals.
Whilst it’s still in the early trial phase within the team, we see it as an essential tool which informs our design decisions, and ultimately enables us to craft outstanding customer experiences.
When writing for Aimee, we make sure to use a conversational tone. It’s important Aimee has an identifiable voice, especially with lots of different designers creating content. We’re developing a tone of voice and persona to improve consistency, add personality and make Aimee more recognisable as EE’s digital assistant.
What’s next for Aimee
Since launching Aimee to the world, it has been an incredible journey of learning and development.
We’re always striving for design maturity through documentation of our processes and principles.
Recent developments in the industry such as Generative AI and Large Language Models (LLMs) opens further opportunity for innovation.
The ambition is to make Aimee our customers’ preferred choice of contact within 5 years. So, there’s definitely room for iteration and continuous improvement. We can draw in brand expression, feature exploration and vision-work to shape Aimee’s future.
Ultimately, our vision for Aimee is to be the most personal, customer-focused digital assistant in the world.
