Short answer: A mortgage database recapture benchmark is useful only when the cohort, denominator, review rules, channel, time period, and attribution window are stated. The right benchmark is not a borrowed percentage. It is a repeatable measurement model that lets an MLO compare approved workflows against the same definitions over time.
This methodology page explains how to create a transparent benchmark from a known borrower or partner database. It does not publish an industry-average conversion rate and does not claim that any particular result is typical. Use actual CRM, communication, calendar, LOS, and finance records for the benchmark your team reports.
Why database recapture needs a methodology
“Recapture” can describe very different activities: a past-client review, a referral-partner touch, a reactivation campaign, a refinance conversation, or a new application from an old record. Those activities should not be combined into one rate without a defined cohort and attribution rule.
A transparent benchmark answers four questions:
Which relationships were selected?
Which records passed review and suppression?
What action counted as a response, appointment, application, or close?
What period and source system established the outcome?
Benchmark definition
Field
Definition
Evidence to retain
Cohort
The relationship group selected for review.
Segment rule, owner, and selection date.
Total records
All records present before filtering.
CRM export count and filter version.
Reviewable records
Records that pass data, relationship, and suppression review.
Review log and exclusion reasons.
Reached records
Records that entered the approved workflow.
Campaign, task, call, or message log.
Unique responses
Distinct responses during the stated window.
CRM activity and channel data.
Qualified appointments
Appointments that meet the team’s written definition.
Calendar and CRM records.
Applications
Applications started in the authoritative system.
LOS or application-system record.
Closed loans
Closed outcomes attributed under the written window.
LOS and finance records.
Use a denominator map, not one headline rate
Report each stage against the denominator that makes sense and label it. For example, a response rate can be responses divided by reached records, while a review rate can be reviewable records divided by total records. A close rate might be closed loans divided by reached records, reviewed records, or a different cohort. All three can be mathematically valid and commercially misleading if the denominator is hidden.
Metric
Example formula
Do not omit
Review rate
Reviewable records / total records
Exclusion and suppression rules.
Reach rate
Reached records / reviewable records
Channel and workflow dates.
Response rate
Unique responses / reached records
Definition of a response.
Appointment rate
Qualified appointments / reached records
Qualification rule and attribution window.
Application rate
Applications / reached records
Authoritative application system.
Close rate
Closed loans / stated cohort
Close date and attribution rule.
Build the benchmark in three stages
Stage 1: Clean and classify
Start with a representative sample. Separate current borrowers, past clients, referral partners, prospects, duplicates, invalid records, and records requiring review. Record the owner, relationship type, last meaningful interaction, known source, and fields that need correction. Do not fill missing facts with guesses.
Stage 2: Run a controlled workflow
Choose one or two segments with enough context to support a useful test. Document the inclusion rule, channel, message purpose, review owner, suppression process, and stop conditions before the workflow starts. AI may assist with summaries, prioritization, or draft preparation, but a human still needs to verify the record context and approved action.
Stage 3: Reconcile outcomes
At 30, 60, and 90 days, reconcile CRM activity with calendar, LOS, communication, and finance records. Report positive responses, appointments, applications, closings, opt-outs, suppressions, data corrections, and time to human follow-up. Preserve the cohort definition so the next test is comparable.
What a useful report includes
Report date, selection date, and attribution window.
Segment definition and total record count.
Review and suppression rules.
Channel and workflow owner.
Stage counts with denominators.
Known data limitations and missing fields.
Opt-outs, corrections, and stopped workflows.
Comparison against a prior cohort only when the definitions match.
Use the worksheet and operating guide
The BNTouch mortgage database recapture benchmark worksheet provides the fields and a labeled synthetic example. The 90-day recapture workflow shows how to turn the measurement model into an operating sequence. A mortgage CRM should help the team see relationship context, assign ownership, apply suppression, and connect activity to outcomes.
For product-fit questions, request a configuration-specific discussion through the BNTouch demo request page. Confirm which fields, integrations, reports, and plan limits apply to the intended workflow before implementation.
Methodology limits
This page does not establish an industry benchmark, promise a result, or replace legal, compliance, accounting, or sales-operations review. A benchmark becomes useful when the team can reproduce its definitions, trace its evidence, and explain what the number does and does not mean.
