Weather Channel AI Experiment
Year: 2018 | Duration: 3 months | Team: Visual designer, Design director, Content strategist
Project manager | Tool: Sketch, AfterEffect
Background
How to using AI to drive more engagement?
Despite being the #1 weather app in the U.S. with over 50 million monthly users, The Weather Channel app has failed to meet IBM’s business goals and user expectations. In recent years, it has lost ground to competitors like Apple’s native weather app, showing a decline in both downloads and engagement.
To boost growth and re-engage users, IBM partnered with AKQA to redesign The Weather Channel app using Watson AI—delivering a smarter, more personalized weather experience.
1.User Journey
I see the weather information, but it’s just numbers to me—I still don’t know how to interpret it
When the project began, I reviewed extensive user interview recordings provided by the IBM team to build a strong understanding of the current user experience. I then focused on a common user journey: checking the weather in advance to plan outdoor activities.
A key pain point that emerged was that users found static weather data unhelpful—it lacked context and personalization, making it difficult to make informed decisions.
2.Ideation
Deep dive into technical Watson AI
This was a major initiative across the entire agency, bringing together ideas from seven different departments. After a thorough review, two concepts were selected to move forward.
As the designer on the project, I took the lead in understanding AI and getting familiar with IBM Watson’s capabilities by exploring demos and taking online courses since AI is still new area back to 2018.
Through in-depth research and team discussions, we identified Watson Visual Recognition and Watson Personality Insights as the most promising technologies to enhance the user experience.
07
Department Involved
76
Ideas submitted
2
weeks brainstorming session
3.Concept 1
Interpret weather in a simple way
We explored using Watson Visual Recognition to create location-based experiences, like recognizing user-submitted photos to give more relevant forecasts.
We also proposed an AR feature that helps users see invisible weather data in a clear, visual way—making it easier to understand and act on.
This could include:
Wind direction
Air quality
UV index
Pollen levels
Humidity
Temperature
And more
This concept was not pursued further because it required significant effort from users to capture weather data themselves, and it depended on devices having high-resolution camera capabilities.
4.Concept 2
Personalized weather forecast
Watson Personality Insights was leveraged to personalize content and notifications based on user preferences and behavior, making the app feel more relevant and engaging over time.
The idea is to design more personalized weather forecast. One that connects you to the weather on a deeper, more meaningful level. It learns from your history, your behavior, your location, your preferences, and gives you advice tailored specifically to you.
I collaborated with the visual design team—who helped redesign the IBM Design System—to finalize the mockups shown below.
5.Outcome
Received High satisfaction from stakeholder
Our client gets inspired by this idea and worked on developing it.
“This is brilliant. It’s reinventing weather apps to be more relevant useful for everyone.”
-Susanna Rodriguez de Tembleque, VP Brand experience of Design, IBM Watson
Year: 2018 | Duration: 3 months | Team: Visual designer, Design director, Content strategist
Project manager | Tool: Sketch, AfterEffect
Background
How to using AI to drive more engagement?
Despite being the #1 weather app in the U.S. with over 50 million monthly users, The Weather Channel app has failed to meet IBM’s business goals and user expectations. In recent years, it has lost ground to competitors like Apple’s native weather app, showing a decline in both downloads and engagement.
To boost growth and re-engage users, IBM partnered with AKQA to redesign The Weather Channel app using Watson AI—delivering a smarter, more personalized weather experience.
1.User Journey
I see the weather information, but it’s just numbers to me—I still don’t know how to interpret it
When the project began, I reviewed extensive user interview recordings provided by the IBM team to build a strong understanding of the current user experience. I then focused on a common user journey: checking the weather in advance to plan outdoor activities.
A key pain point that emerged was that users found static weather data unhelpful—it lacked context and personalization, making it difficult to make informed decisions.
2.Ideation
Deep dive into technical Watson AI
This was a major initiative across the entire agency, bringing together ideas from seven different departments. After a thorough review, two concepts were selected to move forward.
As the designer on the project, I took the lead in understanding AI and getting familiar with IBM Watson’s capabilities by exploring demos and taking online courses since AI is still new area back to 2018.
Through in-depth research and team discussions, we identified Watson Visual Recognition and Watson Personality Insights as the most promising technologies to enhance the user experience.
07
Department Involved
76
Ideas submitted
2
weeks brainstorming session
3.Concept 1
Interpret weather in a simple way
We explored using Watson Visual Recognition to create location-based experiences, like recognizing user-submitted photos to give more relevant forecasts.
We also proposed an AR feature that helps users see invisible weather data in a clear, visual way—making it easier to understand and act on.
This could include:
Wind direction
Air quality
UV index
Pollen levels
Humidity
Temperature
And more
This concept was not pursued further because it required significant effort from users to capture weather data themselves, and it depended on devices having high-resolution camera capabilities.
4.Concept 2
Personalized weather forecast
Watson Personality Insights was leveraged to personalize content and notifications based on user preferences and behavior, making the app feel more relevant and engaging over time.
The idea is to design more personalized weather forecast. One that connects you to the weather on a deeper, more meaningful level. It learns from your history, your behavior, your location, your preferences, and gives you advice tailored specifically to you.
I collaborated with the visual design team—who helped redesign the IBM Design System—to finalize the mockups shown below.
5.Outcome
Received High satisfaction from stakeholder
Our client gets inspired by this idea and worked on developing it.
“This is brilliant. It’s reinventing weather apps to be more relevant useful for everyone.”
-Susanna Rodriguez de Tembleque, VP Brand experience of Design, IBM Watson
