How does Vantage learn my evaluation standard?

Vantage learns your evaluation standard through two complementary channels: one captures what you can clearly articulate, and the other captures the tacit judgment you apply without consciously spelling it out.

Most of the value in your judgment is tacit. You intuitively know which signals matter, which combinations of firmographic data, relationship history, and timing indicate a real opportunity, and which do not. But if you tried to fully document that standard in advance, you would only produce an approximation. The real pattern recognition happens faster than you can consciously describe it.

Vantage is designed around this reality. It does not require you to define your standard upfront. Instead, it observes how you respond to real records and extracts your standard from your actual decisions.

Channel 1: Voice Corpus

The Voice Corpus is created during Phase 1 of setup. A Fulcrum consultant sits with you and walks through real records from your CRM: actual contacts, accounts, and opportunities.

As you review each record, you narrate your thinking:

• Why this one is worth pursuing
• Why that one is not
• What you notice about specific combinations of signals
• What would have to change for you to see an account differently

These sessions generate structured decision-making data. Vantage captures not just the outcome (e.g., pursue vs. pass), but also:

• The signals you pay attention to
• How you weight those signals
• The edge cases you call out
• The exceptions to your own rules

This becomes the foundation of your evaluation standard inside Vantage.

Channel 2: Correction Cadence

Once Vantage is live, the correction cadence captures the tacit knowledge that cannot be fully articulated in advance but shows up in how you respond to real output.

Each week:

1. Vantage produces scored output.
2. You review the results.
3. Wherever the system gets something wrong, you correct it.

Each correction becomes a new training input. Over time, Vantage learns not only the explicit rules you described in the Voice Corpus, but also the more subtle judgment you demonstrate through your ongoing review behavior.

This is why accuracy in Month 6 is substantially higher than in Month 1. The Voice Corpus provides the initial model of your standard; the correction cadence continuously refines it as you interact with real data.

---

Connected questions:

• What is the Voice Corpus and how do I record it?
• What is the correction cadence and why does it matter?
• How long does it take to get Vantage running?

Related: What the Voice Corpus is · The correction cadence · How long setup takes