How the lead reactivation audit works

1. What the audit measures

SignalBack parses the uploaded CSV and maps recognizable name, contact, amount, date, and status fields. Rows without an email, phone number, or estimate amount are unusable. Duplicate contacts are merged during normalization, and the report shows processed, duplicate, invalid, and unknown-status counts separately.

Estimate rows are classified as unsold, won, lost, or unknown. Deterministic status mappings run first. When AI classification is configured, only ambiguous status rows are sent for clarification; rows without enough signal remain unknown and are excluded from the unsold total.

2. Estimate-mode formula

Measured backlog = the sum of estimate amounts classified as unsold. Low planning scenario = measured backlog × 5%. High planning scenario = measured backlog × 12%.

The backlog total comes from the uploaded file. The low and high values are illustrative scenario multipliers, not observed conversion rates for the uploading business.

3. Contact-only formula

If no rows contain estimate amounts, the audit switches to dormant-list mode. Low planning scenario = reachable contacts × 3% × the optional average job value. High planning scenario = reachable contacts × 8% × the optional average job value.

Without an average job value, the report can model a count of potential reactivations but cannot calculate a dollar range.

4. Why these scenario ranges exist

The current 5%–12% estimate range and 3%–8% dormant-list range are product-planning assumptions. They keep reports comparable while SignalBack is in early access; they are not presented as an industry benchmark, forecast, guarantee, or substitute for account-specific results.

When a business has sufficient campaign history, its own eligible-contact, engagement, booking, cancellation, and revenue data should replace generic planning assumptions.

5. What can change the result

• Incorrect, missing, or stale estimate statuses and amounts in the source CRM
• Lead age, project type, seasonality, geography, price changes, and prior follow-up history
• Email deliverability, channel consent, offer quality, response speed, and sales execution
• Jobs booked outside the connected system or contacts represented differently across records
• File-size limits or rows that cannot be confidently classified

Audit outputs are informational estimates, not financial advice or promised revenue. The report is designed to size and prioritize a backlog, not to predict a guaranteed business outcome.

6. Reproducibility and corrections

The same normalized inputs and scenario constants produce the same totals. The methodology date changes when formulas, classification behavior, or assumptions change materially. Factual or methodological corrections can be sent to hello@signalback.ai.