CPLG Survival Rate Dashboard
North star — Survival rate (signed clients who ultimately settle & collect)
—
matured cohorts · goal 40%+
What "survival rate" means and why it's the number that matters
Survival rate = of every client we sign, the share who ultimately reach a settlement we actually collect on. Not signed, not in litigation — paid.
Today it is about 31% on fully matured cohorts, trending to roughly 33% as the last in-process cases resolve. The goal is 40%+.
Why it's the #1 lever: each signed client is worth survival × average fee, about $900. At ~800 sign-ups a month, every 1 point of survival is about $22,000/month in fees, and moving from 33% to 40% is roughly $150,000/month. The biggest place survival is lost is Document Collection — which is exactly where this scorecard focuses.
In Doc Collection now
—
active cases to work
Lost in Doc Collection
—
the leak
Reached demand / settled
—
made it through the gate
Settlement lag
—
retained → settled (median)
The funnel — where matured signed clients end up
Throughput — this period vs the period before
Survival rate by sign-up cohort (vintage) — the trend over time
| Signed cohort | Signed | Settled | Survival | Status |
|---|
Compare matured vintages to see whether survival is improving. Recent vintages are still maturing (cases take ~11 months), so they read low.
Current pipeline — where every case sits right now
Cohort completion — by sign-up month (recent cohorts are still maturing)
| Signed month | Signed | Settled | Lost in doc | In process | Reached demand |
|---|
Read survival off matured rows (≥9 months old). Recent months read low only because cases take ~11 months to settle.
Leading indicators (the early-warning metrics Jose runs weekly)
Welcome-call ≤4hr SLA · % with ≥1 doc by Day 7 · SMS reply rate · RO-obtained rate · cases advanced per specialist. These come from Aloware / SendBlue and the per-specialist queue, wired in next. For now, track them in the weekly huddle against the plan.
5 best ways to increase survival rate — ranked by impact
For the Document Collection Supervisor. Work top-down: #1 unblocks the single biggest reason cases die. ~48 of every 100 signed clients are lost in document collection — every one we save is worth ~$900.
- Get the repair orders and DMV records ourselves — stop making the client chase the dealer.
The full dealership repair-order history is the document that kills the most cases, and dealers slow-walk it. We hand the hardest job to the least-equipped person, and they give up.
- Add DMV, dealer-records, and lender (RISC) authorizations to the retainer packet so we can request records the day a client signs.
- For any case stuck only on repair orders or registration, the docs team requests the records directly from the dealer/DMV under that authorization — don't wait on the client.
- Never tell the dealer it's a claim. Ask only for "the complete service history for this VIN."
- Track RO-obtained rate every week.
- Fix our texting so clients actually see and answer it.
Messages from an unknown number with a bare link read as spam (reply rate ~8%). Text drives most of the documents we collect, so this is the fastest, cheapest lift.
- Send every message from the same number/identity that signed the client (SendBlue), so it lands as a recognized thread.
- Lead with a personal message (client name + their vehicle); send the upload link second, after they reply.
- SMS-first and Spanish-first. One owner of client comms per case — pause parallel messages from intake and other teams so the client isn't overwhelmed into going silent.
- Stop the 9-month linger — decide fast and work winnable cases.
Cases sit ~270 days before anyone advances or drops them, so the team's attention is spread across dead cases instead of the ones that can settle.
- Day 3: fork every new case into Engaged / Needs help / Cold and work them differently.
- Set a forced decision gate at ~30 days: advance it or drop it, no limbo.
- Fix the intake repair-visit count so single-repair cases (which can never qualify) are flagged at intake, not chased for months.
- Run a one-time sweep of the old backlog: one final attempt, then close the unrecoverable cases so the team can focus on live ones.
- Catch bad documents at upload, with a coach — not after days of back-and-forth.
Page-one-only repair orders, expired registrations, and "application for title" instead of the registration trigger endless round-trips, and each one is a fresh chance for the client to ghost.
- At the upload step, show an example of a complete repair order, a current registration, and the RISC.
- When a client uploads the wrong thing, reply instantly and kindly ("this one needs the full multi-page document — here's what it looks like") with a human to help, not a hard rejection.
- Coach clients to request every page of every repair visit, for the same defect.
- Run the team on one number every week, with per-specialist accountability.
The process already exists; the gap is execution. Make daily progress visible so nothing quietly stalls.
- The one number: cases advanced to Supervisor Direction per new sign-up cohort — not messages sent or calls made.
- Give each specialist a weekly scorecard: welcome-call ≤4hr hit rate, % of their cases with a document by Day 7, and cases advanced.
- Make Franchesca a second approver to Supervisor Direction so completed cases never wait on one person.
- Audit the CNMR bucket weekly so hard cases can't be parked there to hide.
Source: HubSpot deals (live). Survival and lag refresh automatically. Questions → Sam.