Advisor: AI-driven guided analytics

The design process and system

I helped develop the design process we use at Adthena as an iterative five-stage approach. I maintain the Design System Style Guide in Sketch, to ensure consistency throughout all areas of the application.

Adthena's five-stage iterative UX design process. — The five-stage design process and the design system style guide.

The Adthena design system style guide in Sketch. — The five-stage design process and the design system style guide.

Mapping the user journey

I identified the user journey phases a customer moves through, based on the user interviews we had previously conducted. I then mapped out the proposed user journey from interviews with Customer Success Managers and stakeholders, to be validated through user research. I also aligned this with our technical architects to get validation of what was technically possible.

A detailed onboarding service blueprint mapping emails, activities, pain points and feature ideas. — Onboarding journey mapping: activities, pain points and feature ideas.

The proposed Advisor user journey mapped into five phases. — The proposed Advisor journey, mapped into five phases.

The research plan

The next step was to create a research plan. I mapped our current learnings from customers to identify what our research questions were, and the corresponding questions we would ask in the subsequent user tests.

Research questions mapped across the five journey phases. — Learnings mapped to research questions across the journey.

Building the Advisor alpha

We have a section of the app called 'Labs' where we release very early work in progress for user feedback. I worked directly with our lead technical architect to create an 'Alpha' version of the guided analytics app called Advisor, which was the basis of what we would be putting in front of users to gather user research.

The Advisor alpha built in the Labs section of the app. — The Advisor alpha, released through Labs for early feedback.

The Advisor prototype. — The Advisor alpha, released through Labs for early feedback.

Analysing the research

Remote interviews were conducted and recorded over Zoom with the help of our in-house User Researcher. I worked with him to analyse the interviews in Dovetail, which provides automatic transcription of the user tests and the ability to tag and organise feedback. We have 3 letters of intent signed by customers to use the final 'Advisor' feature, and predict a significant increase in potential client revenues based on an entirely new function we will be able to offer.

Analysing user interviews in Dovetail, with tagged themes. — Interview analysis in Dovetail: emerging themes, solution themes and technical exploration.

Closing the feedback loop

For the 'Labs' section of the app, I created a banner to make it easier for customers to give us feedback. Below is the reference spec I created in Sketch and uploaded to Zeplin for the developers to reference. After this was released it resulted in a 30% increase in monthly customer feedback, which in turn meant I could work with the Product Owner to prioritise future user-centric improvements.