Faster Tour Search: Designing for Learnability

The Approach

1. Defining success across teams—speed and trainability
   Prior work had optimized for features, not learnability. I facilitated a workshop to align on both.

2. Reframing search as a decision-making system
   I led the design of a unified workflow that brought key booking criteria into one place.

3. Proving value before committing to build
   Validated with agents that the new flow improved speed and confidence (Rated 9/10).

My Role

• Aligned product and business stakeholders on core agent decisions—not features

• Defined a decision model for booking tours alongside engineering: Find → Compare → Evaluate

• Set learnability as a constraint, prioritizing onboarding speed over completeness by co-creating the roadmap with stakeholders

• Led high-fidelity design execution and partnered with another designer on research and testing facilitation.

The Challenge

Four weeks to onboard was too long. The goal: cut it to two without sacrificing accuracy or speed.

Agents had to evaluate tours across pricing, availability, and constraints—but the system worked against them.

The costs

Missed opportunities, lower conversion, and longer training times to navigate the system.

Defining the Problem

What wasn’t working

Past work was misaligned on both outcomes and target users.

I led a cross-functional workshop where stakeholders voted on success metrics and core personas—aligning on who we were designing for before defining solutions.

Success metrics:

• Reducing training time by 50%

• Reducing call time by 30%

• Reducing travel partner call center reliance

Our core user became tour agents—over outbound sales and third-party agents—sharpening scope for the roadmap.

Onboarding time + benefit to tour agents became the force function to resolve competing stakeholder priorities.

Roadmapping: Building the System in Learnable Stages

I synthesized workshop outcomes into a shared roadmap across product, engineering, and business.

Product pushed for advanced filtering upfront, but I prioritized time to first booking over configurability to optimize onboarding speed.

I advocated that we lead with search: a familiar entry point that lowered training burden, especially as agents were defaulting to the consumer site.

I prioritized a sequenced decision model—trading flexibility for faster, more confident decisions under time pressure.

This flow optimized learnability, reducing training complexity while we built toward a unified system.

Key Tradeoffs

• Prioritized decision speed over full configurability
  Focused on a clear search entry point instead of exposing all filtering logic upfront.

• Leveraged existing mental models over introducing new ones
  Aligned search with the consumer experience agents were already using.

• Reduced initial scope to improve learnability
  Deferred advanced features until the core workflow was established.

UX Research

Research Findings: Too Little, Too Late

Talking with agents revealed pain points:

1. Key information not accessibile during live calls → low customer confidence and missed bookings

2. Workflows didn’t support real-time decisions → reliance on memory and workarounds

3. Information spread across 5 sources → time pressure doing manual synthesis across agent tools, website + printouts

Wireframing & Prototyping

Designing for familiarity

Agents often relied on the consumer website to search because it was more intuitive than internal tools.

The new design leveraged existing patterns from the website to reduce the need to relearn how to search.

The After: Search + Results in One Place

The search bar is now layered on top of the legacy interface as an overlay—creating a unified entry point without restructuring the system.

I negotiated with engineering to land on this approach given the constraints—trading a full UI overhaul for a faster, lower-risk path to unify search and results.

This enabled:

• Search + comparison in one place

• Continuity & integration with existing workflows

The Before: Search + Results Separated

Agents had to switch contexts to compare and evaluate tours.

Now, they’re in a single interface—enabling agents to search, compare, and decide without breaking flow.

Key Reflections

• Aligning on shared goals early prevents rework down the line, keeping teams focused on outcomes

• Validating value before build reduces risk and ensures solutions actually improve real workflows

• Designing for learnability is critical in training-heavy environments