Last Updated: January 2025

TL;DR

OneSource Cloud used Aurium to rapidly test multiple ICP (Ideal Customer Profile) hypotheses through LinkedIn outreach. The platform's conversational AI and reinforcement learning enabled quick iteration—learning which prospect segments responded best in weeks rather than months.

The Challenge: ICP Uncertainty

Multiple potential customer segments. Limited resources to pursue all simultaneously. Needed data on which segments convert best. Traditional testing methods too slow.

The Solution

"Test multiple ICP hypotheses really quickly"

  1. Define multiple ICP hypotheses
  2. Run parallel outreach to each segment
  3. Measure conversation and conversion outcomes
  4. Let reinforcement learning identify patterns
  5. Adjust targeting based on data

The Results

Speed of Learning:

  • Traditional timeline: 3-6 months per hypothesis
  • With Aurium: Meaningful data in 4-6 weeks

Key Insights Gained: Identified top-performing industry verticals, discovered optimal company size range, pinpointed most responsive titles, eliminated low-converting segments.