Tabnam- Conversational AI for User Feedback
Tabnam uses an AI chatbot to collect feedback conversationally via text, WhatsApp, and more, while aggregating and analyzing data from platforms like Discord and Zendesk.
ROLE
DURATION
TEAM
Founding UX Designer
4 months
3 engineers, product director, PM
PROJECT ACCOMPLISHMENT
Designed Tabnam’s AI-Powered Product Suggestions from client research, driving 20+ new clients and major growth.
PROBLEM
Many businesses struggle to collect meaningful and actionable customer feedback. Traditional surveys often lead to low engagement and incomplete insights.
How might we help businesses collect and analyze actionable feedback to improve their products?
SOLUTION
An AI-powered dashboard designed to collect valuable customer feedback through conversational AI, providing businesses with a platform to analyze the gathered data.
PROCESS
To tackle this problem, I followed the steps outlined below. These steps weren't linear, and I iterated as needed, going back and forth as required.
Choose your desired cleaning service
Research
Use the Cleanic App for laundry pickup
Ideation
Leave us the keys and we'll take care of the rest
Design
Welcome back to your home clean home!
Test
RESEARCH
Reviewed previous customer data & Competetive Analysis
I began by delving into the existing data on client wants and needs, as well as insights gained from previous client interactions. This process helped me grasp what had already been developed and identify areas for further improvement. Working closely with the CEO, we analyzed these findings to uncover common themes in client requests and preferences, which informed our strategies for future development.
After reviewing previous data, I conducted a brief competitive analysis of other feedback management systems in the market to understand industry standards and identify areas for improvement.
What was business saying
More customer insights in less time
RESEARCH FINDING
These themes consistently emerged from customer emails, in-person conversations, and phone calls, as well as through examining other potential solutions in the market.
Overwelming amout of data
Many customers expressed that they find the amount of data overwhelming and prefer it to be categorized into distinct sections.
Security and Privacy Concerns
Customers expressed a desire for heightened security and privacy measures to protect sensitive or confidential information displayed on the dashboards
Data Customization
Customers expressed the need for increased customization options on the dashboard to cater to their specific requirements, like filtering through feedback or displaying only negative feedback to certain users.
Data Visualization
Certain customers requested enhanced data visualization, such as graphs and illustrations, to facilitate better comprehension of the data.
IDEATION
Creating a new process infused with AI to increase responsiveness and reduce human work
After mapping out the entire feedback process that business owners follow, I identified key steps where AI could automate and streamline tasks typically handled manually. This led to the development of the Conversational AI chatbot.
From there, I created a rough journey map to highlight gaps and areas where new user flows and interfaces needed to be designed.
Site Map
Brainstormed and came up with the site map to understand the user flow and what the design needs to account for. I divided it into four sub pages: Dashboard, Feedback, AI Chat, and Settings.
I ideated various layout options to find the most efficient way to display information on the dashboard, the first page users see after logging in. To determine the most critical data, I collaborated with the team and gathered user research to prioritize users’ needs and preferences.
I brainstormed and came up with a wireframe for each key page. Specifically, I have categorized each insight section on type of complaints/praises the business would receive. Instead of generic titles like "Top Complaints", I made it into more personable recommendation system. For ex: for a restaurant business, the feedback card titles AI would process is "Favorite Ingredients", "Interior Improvements", etc.
TESTING
TESTING
In order to identify any design flaws, ensure functionality, and gather diverse user perspectives on the design, testing was conducted through interviews with users who interacted with the dashboard and chatbot. Here is the data we obtained from these interviews.
For the final testing, I worked with five clients to gather insights based on three key principles:
-
Useful: Can customers easily accomplish their tasks using the dashboard? Are all necessary features included and effortless to use?
-
Learnable: Is the dashboard intuitive enough for clients to navigate without training?
-
Reasonable: Can tasks be completed within a reasonable amount of time and effort?
Here are the insights I gathered from testing:
“I failed to discover or engage with the conversational AI feature due to difficulties in locating it. it wasn't easy to figure out. "
Client#1
"I have to take screenshots of the dashboard to share it with my team, which is not super convinient method of sharing. It is recommended to implement an easier way to share the dashboard with the team."
Client#2
“I love the The feedback section of the dashboard. It helps me to quickly see the user input and take actions apon it.
Client#3
INTERACTIVE PROTOTYPE
After iterating and refining the dashboard based on testing feedback, I moved on to finalizing the designs.
Dashboard homepage becomes the first page users land on after they log in. They can easily navigate between 4 sections to find what they need.
This new dashboard is highly received and loved by our users.
FINAL DEISGN
Profile & Product
I set up a dedicated tab for our products, allowing customers to select which ones they want to display on their dashboard and specify the insights they're interested in. Additionally, the profile tab provides details on the account holder and enables them to edit their information.
User feedback
Featuring various subscription plans and option to change it.
Dashboard homepage
The dashboard allows clients to select one of their products and view both positive and negative reviews, along with the product's overall score and recommended actions. This feature empowers clients to address the question of 'how can we improve our product to meet our customers' needs.' By leveraging these insights and organizing customer feedback, clients can enhance their products and make more informed business decisions.
Prompting questions
If clients wish to delve deeper into a specific insight, they can access comprehensive details by clicking on the review table directly from the dashboard.
Mobile Design
WEBSITE DESIGN
I completely redesigned the Tabnam website and spearheaded the creation of their entire branding from the ground up. This marks the first iteration of the brand's transformation.
LEARNING
Data Visualization for AI
I learned how to design intuitive dashboards that effectively present AI-analyzed customer data, ensuring complex insights are easily digestible for users
Cross-Functional Collaboration
I employed design workshops with engineers and product managers to align on AI functionality and ensure feasibility, reducing miscommunication and speeding up implementation.
Structuring Data Hierarchies:
card sorting helped me to organize customer data into clear categories and prioritize key insights, ensuring the dashboard displayed the most critical information first.
Fast-Paced Startup Environment
Working at Tabnam taught me to adapt in a fast-paced startup environment, balancing research, design, and testing while delivering results with limited resources. Collaborating with a small team strengthened my communication and rapid iteration skills, ensuring impactful, outcome-driven designs.