UX Leadership: Visualisation makes the Strongest Cases

Audience preference visualisation

Labeling and Preference Categorization are the two tools I always use at the requirement phrase and the feedback phrase of a project.

This pie chart example showcases the labeled user feedbacks that are gathered from the chat ad conversation.

* The design of these "topic names" is aligned with the keywords(campaign focus) on the client website.

‍

‍

‍

‍

‍

Keywords analysis

Response relevancy is an important metric in Chatbot UX. Knowing the right keywords helps development team to train the chatbot and provide more relevant answers.

This bar chart example marks my effort to find out what keywords are mostly used in the open-inputs from the audience so that we can further determine how many response variations we need for a similar topic.

‍

‍

‍

‍

‍

‍

Version control & Performance analysis

By means of control variates, we can find out what UI elements, layout design, colour, tagline and sub-title, button amount, button size, font weight, font size generate better campaign performance.

Measuring UI quality based on marketing metrics gives stakeholders guidance beyond a single campaign.

Check out this example for more details about controlled testing different UI designs.

‍

‍

‍

‍

Data-driven insights

Any chatbot ad creation will face with the following challenges:

• Conversation needs to reflect brand website structure and content
• Conversation needs to be short and engaging
• Feedback & d

‍