BBB AI Assistant (Beta)

Helping consumers get to the right BBB action faster and with confidence

Role: UX/UI Designer
Platform: BBB.org (consumer entry experience)
Status: Live (Beta)
Timeline: 2025 (Beta)
Deliverables: Conversation UX + content, dialog flows, interaction design, high-fidelity prototypes, usability validation

Impact statement: A lightweight AI-guided “traffic cop” that routes users to the correct path: File a Complaint, Leave a Review, or Report a Scam, while keeping a clear manual fallback (“Standard Mode”) for full user control.

Project Overview

BBB supports consumers who arrive with a problem but aren’t sure which BBB pathway fits their situation. At the “How can we help?” entry point, users may be seeking a resolution (complaint), sharing feedback (review), or reporting a scam (scam report). When users choose the wrong path, they lose time, abandon, or submit incomplete reports, creating extra friction for both consumers and support teams.

The AI Assistant (Beta) was introduced to let users describe their issue in their own words and be guided to the correct destination, with the ability to switch to a non-AI “Standard Mode” at any time.

Problem Statement

Research identified four primary UX issues:

ProblemImpact
Users struggle to decide between Complaint, Review, and ScamSubmitted claims end up in the wrong place, increasing the need for manual work for BBBs
Entry experiences require scanning definitions and making a high-stakes choice upfrontUsers hesitate, second-guess, or abandon before starting
Sensitive information risk in open-text descriptionsNeeded clear guardrails (do not include SSN or full card numbers)
Some disputes require specialized handling Needed clear guardrails (ex, do not include SSN or full card numbers)

“Let’s get you to the right place!”

Key UX insights

  1. Users want to explain the situation first, then be guided, rather than blindly choose a category.
  2. Trust increases when the experience offers a transparent fallback (manual “Standard Mode”).
  3. Clear safety language reduces hesitation and prevents oversharing sensitive info.
  4. Example prompts reduce friction in blank states and help users quickly write a useful summary.

Design Goals

  1. Reduce time to reach the correct BBB destination (Complaint, Review, or Scam Report).
  2. Increase confidence at the moment of choice by guiding users from their own language.
  3. Preserve user control with an always-available manual path (“Standard Mode”).
  4. Add safety guardrails to discourage sensitive info entry and set expectations
  5. Handle edge cases with clear routing
AI Assistant (BETA)
Standard Mode

Validation & Feedback

MeasureStandard entry choicesAI Assistant (Beta)
Confidence in “I’m in the right place.”Mixed, relies on interpretationHigher, guided by user language and manual fallback
Steps to reach the right workflowCan require backtrackingMore direct routing after one prompt
Comfort sharing detailsRisk of oversharingImproved with visible guardrails and safety copy