e-Comm website & chatbot for the university research
Background
Customer Request
.png){kind=logo}
Challenges & Solutions
Challenges
Need to recreate a real-life e-Comm experience for participants
For her research, the customer needed to recreate the experience of a real-life e-comm journey to study participants' behavior and choices in prechosen circumstances.
Study user behavior patterns with store chatbots
The focus of the research was to study the users’ interactions with chatbots that provide suboptimal recommendations during online shopping. So, the customer's challenge was to create chatbots that would meet research requirements.
Analysis and management of research data received
During the research, it was crucial to store all the received data in one place, organize it neatly, and keep the data precise and accurate for providing explicit research results.
Solutions
e-Commerce website development
The FTL team built an e-commerce website that makes the impression of a real headphone store.It didn’t include all purchase flows but focused on simulating realistic communication with machine- & human-like chatbots for research purposes.
Machine-like and human-like chatbots
We built various machine- & human-like chatbots with different typing algorithms and AI avatars to assist users during a simulated shopping experience. It helped study how chatbot communication influences user behavior during purchase.
2 databases to store and process the information gathered
We built two Notion databases: one – for storing structured user responses and another – for individual chatbot conversations. This setup made it easy to track interactions, analyze behavior, and debug a website when needed.
.png)
project facts
main features
Participant Onboarding
We've created onboarding pages on the website to ensure the participants have a smooth user experience and impression of real-life shopping journey during their purchase.
Machine- and human-like chatbots with custom logic
We implemented different chatbot types — from a basic machine-like bot to more human-like ones, including a talking AI avatar. Each followed a predefined logic for helping users choose the most suitable headphones, designed by the client.
Research databases
The databases integrate and store all participant information automatically and have a smart system of fields, filters, and requisites to provide all necessary data for ongoing research and retrieve it efficiently.
Technologies used
Customer Review
I like the way we worked with Faster Than Light, their communication and efforts brought to understand my needs and goals. Although we might have done estimations and planning of the project a little bit differently, I guess it is experience and insights you get only while diving deep into the project and its specifics.