Glossary
The LO Radar vocabulary.
Plain-English definitions for every term we use across the platform.
- Pipeline NPV
- The present value of a mortgage LO's career book of past clients, priced as an asset that moves daily.
- Drift Radar
- Per-borrower churn prediction scored against each client's personal contact rhythm — not a generic threshold.
- Voice Training
- A system that learns the LO's writing voice from past sent messages and renders every draft in their tone.
- Rate-Drop Radar
- Past borrowers ranked by projected annual savings if they refinanced today at current market rates.
- ARM Reset Calendar
- Every 5/1, 7/1, and 10/1 ARM in the LO's book, mapped to its reset date with a 12-month pre-reset cadence.
- Anniversary Triggers
- Auto-drafted outreach at year 1, 3, 5, 7, and 10 of a past borrower's purchase. Year 5 is statistically the highest-conversion refinance year.
- HELOC Radar
- Equity-rich past borrowers worth a HELOC conversation. Sorted by available equity × rate spread.
- Cash-Out Detection
- Borrowers whose appreciated equity could fund a renovation, second property, or business — flagged by life-event signals.
- Second-Home Signals
- Lifestyle indicators (vacation searches, kid heading to college, job relocation) that suggest a second-home purchase is imminent.
- Credit-Improvement Flags
- Past borrowers whose credit score has improved 40+ points since origination — they may now qualify for a meaningful rate reduction.
- RESPA
- Real Estate Settlement Procedures Act. Federal law restricting kickbacks and disclosure requirements in real-estate transactions.
- TILA
- Truth in Lending Act. Requires lenders to disclose loan terms clearly and uniformly across products.
- TCPA
- Telephone Consumer Protection Act. Restricts automated phone and SMS contact without prior consent.
- LOS
- Loan Origination System — the core software an LO uses to originate, process, and close mortgages. Encompass, Calyx, Byte, MeridianLink, LendingPad.
- Empirical Bayes
- Statistical technique that blends a borrower's personal contact history with a population-level prior — produces stable cadence estimates even with thin data.