Overview
In 2024, GoSky Chatbot Platform faced several business challenges, including a decline in Facebook business usage and low subscription rates. Based on market research and user studies, we decided to expand into two new channels (Threads, Instagram) and improve usability.
As a result, the platform successfully launched both new channels and increased subscription rates through enhanced user experience.
My role
Improve product usability through UI/UX design, increasing the subscription rate.
Collaborate with the manager and PMs to align on project goals and define success metrics.
Analyze product data and conducted 5 user interviews and 3 usability tests.
Iterate on the design system to align with the product positioning.
Expand into 2 new social channels (Threads, Instagram).
About GoSky Chatbot Platform
GoSky Chatbot Platform is a CRM dashboard designed for marketers, with active users in Taiwan, UK, US, and HK. The system integrates with Meta's API services to automate interactions on social medias, including comment replies, private message responses, message broadcasting, fan profile analysis.
Problem & Challenge
Based on our market research and user studies, we identified the following key problems:
Marketers and content creators faced numerous obstacles during setup, including frequent configuration failures and uncertainty about whether settings were applied successfully.
The average subscription rate was only 0.3%, which remained low despite efforts to expand into overseas markets. This not only impacted revenue but also indirectly led to a decline in external investment.
To differentiate ourselves from competitors, the product was positioned as a self-serving platform. However, the initial designer had directly copied competitor designs, resulting in a mismatch with our target audience and the accumulation of significant design debt.
Research
I reviewed product data, participated in 5 interviews, and conducted 3 usability tests. Through this process, I identified the main product issues:
The highest user drop-off occurred during the "Connect Account → Publish" flow.
The setup process was overly complicated, with too many unnecessary fields.
Users struggled to understand whether they had completed the setup successfully.
Design process
By integrating the setup features directly into the Canvas, we reduced page switching and decreased user drop-off.
We addressed feature logic from the IA. Interviews revealed that fewer than 10% of advanced users utilized the "multi-condition" feature. Based on the product’s positioning, we simplified the setup logic into 2 layers.
Test & Iteration
After testing the 1st version, I observed that 60% users repeatedly clicked when setting the "multi-item sync settings" feature. The issues included:
Unclear clickable area and visual hierarchy
Information flow didn't match user expectations
Then I proposed 7 design variations, discussing usability with the PM, confirming technical feasibility with the RD team. Finally moved forward with a 2nd round of testing.
Here's the final version of blocks design in Canvas:
Design iteration for product positioning (self-serving)
To better align the product design with its positioning (self-serving), I conducted multiple iterations focused on the consumerization of B2B.
Preview pages: I designed preview pages that allows users to visualize their setup from end-users' perspective. This helped bridge the gap between 2B and 2C users.
Aha! moment: During testing, I found that users experienced a high level of uncertainty during the setup process, often questioning whether they had completed the configuration successfully. I enhanced the emotional design to reinforce user confidence and satisfaction. And finally increased the “perceived completion of publishing” rate from 66% to 100%.
Pricing strategy: To better match the behavior of 2C-style users like creators and KOLs, I replaced time-limited trials with usage-based limits, clearly shown on the homepage. This reframed restrictions in a positive way, reducing friction and improving trial conversion.
Hand-off
UI Flow & UI Spec
Design System & Component Guideline
Result & Impact
This project led to several impactful outcomes and positive feedback.
Reflection
Due to time constraints, this project adopted a growth design approach. I leveraged data-driven decision-making and conducted multiple A/B tests to help the team achieve our project goals. The final outcome delivered positive results and meaningful business value for the team.
What I brought to the team…
Persuading the team by applying different design strategies to accomplish the project goals.
Translating business problems into design solutions that addressed real user needs.
Collaborating closely with team members, using rapid testing and iteration to find user-friendly designs.
