The AA chatbot and help centre tools

Role: Lead UX Writer

Brief: Minimise inbound customer care correspondence by enhancing the help centre offering

Responsibilities: Digital strategy, data analysis, tech development, concepting, UX, content design, copywriting, UI


As part of a wider digital transformation, we were tasked with making customer care for over 10 million users more efficient and alleviating the load on the Customer Service agents. Unfortunately the company was going through hard times - over half of the team were being made redundant - so the team were having an even harder time keeping on top of the 3,000 daily calls and emails.

So after extensive customer research, we decided to develop 2 overlapping solutions:

  • An exhaustative self-service portal for customers to find the information they need online

  • A chatbot for customers who feel more comfortable getting this information by engaging in conversation


Self-Service Portal

The existing experience was confusing and the content was incomplete or led to dead ends. Customers had to call or email to resolve their query.

To fix this we created a fully end-to-end, informative and actionable tool that covered all likely customer queries across all 5 business divisions. Gathering and analysing this data was no mean feat, but our comprehensive background work had to appear clean and clear to the user.

Our UX concept was to develop rule-based interactions that put the customer in control of their own personal journey - from specific query to specific answer. First, query scenarios are split into 11 topics then drilled down further into known queries within that topic.

Finally, we give customers the answers they need or, if that’s not possible, more than one next step to resolve their query. Our card-based UI allowed us to do this with visual consistency, culminating in 4 types of ‘answer’ card - onscreen information, online task, recommended reading and contact details.

Chatbot

As an alternative to the self-service portal, the chatbot follows a similar UX concept but with added artificial intelligence and intuition so customers can communicate their way to the relevant answers or information they need.

It offers multiple ways to interact, from a free text chat box to selection buttons, again giving customers the opportunity to interact in the way they’re most comfortable.

It also includes extra levels of personalisation and efficiency for logged-in users including name recognition, product subscription knowledge and Customer Service correspondence history.

As well being behind the UX, UI and tech thinking for both tools, I also wrote the conversations and interactions - turning a huge data matrix into inclusive, professional, informative, clear and decisive content.

In the first month after launch, inbound calls dropped 30%. Job done.

Please have a play around yourself.