AI Chatbot for Jacklyn De Ciccio

An AI-powered chatbot for a Toronto real estate broker — handling off-hours inquiries, recommendations, and scheduling.
UX/UI Design
UX Research
Product Design
Role
UX/UI Researcher, UX Strategist
Client
Jacklyn De Ciccio
Duration
6 weeks
Status
Shipped
Link
AI Chatbot hero

Background

Partnership with a Greater Toronto Area real estate broker to develop an AI-powered chatbot addressing client communication gaps during off-hours browsing.

Problem — Communication bottlenecks caused delayed responses to potential clients. Peak browsing occurred outside standard business hours, creating missed engagement opportunities.

Solution — AI-driven system offering instant inquiry responses across Instagram and website, personalized property recommendations, and real-time calendar-synced appointment scheduling.

Result — Breaking the appointment scheduling process into simpler steps increased successful bookings by 40%.

UX Approach

Researched the broker's workflow through interviews and observation. Found 70% of inquiries arrived via Instagram and website after 6 PM — outside business hours.

Key insight — Clients needed a response in minutes. The chatbot had to feel like the broker was present, not like a FAQ bot.

Process

Heading

Context

The Communication Gap

The Communication Gap

Image caption

Paragraph

Real estate clients browse listings after hours — on Instagram, on the website, late at night. But brokers can only respond during business hours. Every unanswered message is a lost lead.

Real estate clients browse listings after hours — on Instagram, on the website, late at night. But brokers can only respond during business hours. Every unanswered message is a lost lead.

Image caption

Stats

70%

of inquiries arrive after 6 PM

+40%

increase in completed bookings

3 steps

booking flow, down from 7

Image caption

Callout

The chatbot needed to feel like the broker was present — not like a FAQ bot. Tone drove trust more than features did.

The chatbot needed to feel like the broker was present — not like a FAQ bot. Tone drove trust more than features did.

Image caption

Heading

Design

Conversation Design

Conversation Design

Image caption

Annotated Image

Chatbot conversation flow wireframe

Image caption

Chatbot conversation flow wireframe

[{"x":20,"y":30,"label":"Quick reply chips reduce typing friction"},{"x":65,"y":55,"label":"Availability badge sets expectation upfront"},{"x":40,"y":78,"label":"Booking confirmation previewed inline"}]

Two-Column Comparison

Image caption

V1 — Open-ended chat

Users could ask anything. Most typed nothing — blank input paralysed 60% of first-time visitors.

V2 — Guided prompts

Three starter questions surfaced on load. Completion rate jumped from 38% to 71% in the first week.

Pull Quote

It felt like I was there even when I wasn't. Three bookings came in overnight.

Jacklyn de Ciccio, broker

Image caption

Stats

+40%

increase in completed bookings

3 steps

booking flow, down from 7

70%

inquiries arrive after 6 PM

Image caption

Design Impact

Simplifying the booking flow from 7 steps to 3 dramatically increased completion in testing.

+40% increase in successful bookings

Seconds response time, down from hours

24/7 coverage across Instagram and website

Takeaways

AI tools work best when they feel human. The chatbot's tone, not its feature list, drove trust in testing. Brevity and confidence in replies outperformed thoroughness every time.

Reflection