Daniella's Customer Brain

Honeycomb <> OneAI: Weekly Sync

3 min read

Honeycomb <> OneAI: Weekly Sync

Date: 2026-01-27 | Duration: 26.579999923706055 min | Account: honeycomb

External attendees: jan.j@honeycombinsurance.com, jon@honeycombinsurance.com, rafael@honeycombinsurance.com

Summary

The team addressed the significant increase in lead disqualifications, which disrupted operations and caused churn, prompting Rafael Mintz to pause lead ingestion on his side. Amit Ben agreed to conduct a thorough review and develop benchmark tests to resolve logic errors causing excessive disqualifications. To improve operational efficiency, Rafael and Jon emphasized refining the disqualification logic and clarifying pricing communication to avoid ambiguity. They proposed adding a read-back step during calls to confirm prospects’ intentions and disabling disqualification-triggered file closures to keep leads active. The introduction of an automated QA benchmark system was also discussed to ensure consistent AI behavior prior to deployment, with strategic goals set to stabilize AI qualification and reduce operational drag for new use cases.

Key Points

🚫 Lead Ingestion Paused: Team paused lead ingestion due to rising disqualifications causing operational disruptions and increased churn. 🔍 Error Review Commitment: A thorough review of disqualifications will be conducted to identify logic errors costing business. 💬 Pricing Communication Clarity: Agreed to differentiate between non-binding price indications and formal quotes to avoid prospect confusion. 📂 File Closure Adjustments: Disqualification-triggered closures will be disabled to keep leads active for final decisions. 🤖 AI QA System Implementation: An automated QA benchmark will run tests before deployment to ensure consistent AI behavior. 🎯 Strategic Goals Set: Aiming to improve disqualification logic and pricing communication while reducing operational drag for new use cases.

Overview

  • Lead Ingestion Paused: Team paused lead ingestion due to rising disqualifications causing operational disruptions and increased churn.
  • Error Review Commitment: A thorough review of disqualifications will be conducted to identify logic errors costing business.
  • Pricing Communication Clarity: Agreed to differentiate between non-binding price indications and formal quotes to avoid prospect confusion.
  • File Closure Adjustments: Disqualification-triggered closures will be disabled to keep leads active for final decisions.
  • AI QA System Implementation: An automated QA benchmark will run tests before deployment to ensure consistent AI behavior.
  • Strategic Goals Set: Aiming to improve disqualification logic and pricing communication while reducing operational drag for new use cases.

Action Items

Amit Ben Complete thorough review of disqualification calls and create benchmark tests to simulate edge cases and verify fixes (03:33) Develop explicit categorization for disqualification outcomes and calibrate auditing system (11:15) Suggest and provide precise phrasing for call flow language clarifying ‘non-binding price indication’ vs ‘quote’ before deployment (18:28) Incorporate a ‘read back’ confirmation step after disqualification to reduce false positives (22:18)

Jon Repka Provide example language clarifying ‘indication’ vs ‘quote’ to OneAI for call flow refinement (13:47) Resume lead transfer but disable Honeycomb-side disqualification closing actions until updated logic is validated (18:38)

Rafael Mintz Stop triggering closing actions on disqualifications internally pending resolution to prevent churn and business disruption (17:00) Review specific pricing indication cases and prepare to discuss solutions with OneAI (18:50) Collaborate with Jon and Amit on finalizing language clarifications and logic flows related to pricing and quotes (18:50)

Graph View