AI design sprint

Designed AI feature concepts to streamline property listing on Sykes’ Owner App

My Role: UX Designer

Squad: 2 UX Designers • UX Research • 3 POs

Tools: Figma • Userzoom • Lovable AI

Duration: 1 week

Business challenge

This design sprint allowed exploration into how AI could shape the future of Sykes' app products. My group created two feature concepts to improve the Owner App onboarding experience.

Approach

  1. Identified pain points in the existing user journey.

  2. Ideated and designed two app feature concepts.

  3. UX research conducted moderated usability tests with property owners, enabling feedback and iterations.

Research review workshop

Collaborative review of existing user research and personas revealed key pain points:

  • Onboarding Time: New owners took an average of 232 days to go live. Airbnb allows instant go-live.

  • Limited Availability: Owner support only available 9-5 via phone.

  • Inefficient Communication: Emails delayed contract signing.

  • No Listing Preview: Owners couldn’t preview listings, affecting satisfaction and bookings.

“I started my onboarding journey with Sykes in March 2024, and my property has only just been listed today.”

-Owner, November 2024

App feature ideation

We held a crazy 8s session to generate ideas around streamlining owner onboarding while leveraging AI. Concepts were prioritised through team voting and discussion.

Key questions:

  • How might we give users control over onboarding?

  • How might we improve description transparency?

Final Concepts

Concept 1: Property Listing Builder

AI gathers all available details for a property to provide a listing preview, using just an address.

Concept 2: AI Smart Companion

AI-powered image recognition to identify features in a property, speeding up the listing process.

Concept 1: Property Listing Builder

Get started - Once the user has verified their address using ID, they are given an overview of the property listing journey.

Find property - AI can gather key location information and property details from the address, saving users time and effort.

Create listing preview – Users stay informed as AI gathers data from multiple sources.

Preview listing – Data gathered by AI is auto-populated into a draft listing, with any missing information highlighted. Users with listings on other platforms likely to see a more complete draft.

Concept 2: AI Smart Companion

Finalise listing - AI image recognition identifies property features and guides owners through optimising their listing.

Complete listing - Owners review the AI-generated listing and confirm all details before confirming, ensuring it meets their expectations.

User research tested both feature concepts

Throughout the 1-week sprint, UX Research ran moderated usability tests of mid-fidelity concepts. Participants were existing property owners, and their feedback fed into iterations before the presentation.

Both concepts explored ways to reduce friction in the onboarding journey, while maintaining user trust and control. Initial feedback is positive, and refinement will take place under a future initiative.

Key insights:

  • First-time listers were pleased with earnings estimates.

  • Some users who weren’t as comfortable using technology, preferring human support. Clear explanations of AI data usage built trust.

Reflections

Next steps

  • Evaluate technical feasibility with developers.

  • Share insights with other squads for visibility.

  • Further refine concepts based on initial feedback.

  • A great opportunity to collaborate across squads to explore how AI can be leveraged to streamline the onboarding process.

  • We presented the final concepts to senior leadership, who chose to move our ideas forward in a future initiative.

↑