Retention ConsoleSee churn in week 3,
not month 6.
Supply-side retention and finance intelligence for VETTX: every dealer account health-scored against its market's real yield, a daily run sheet that says who to call and what to say, and a cash forecast that prices churn before the renewal call does.
Every number on every screen is a live Postgres query — filter questions are safe.
Units bought · MTD
1,847
Offers sent today
312
Listings scanned
48,112
Revenue saved
$63.4k
Account health · Riverbend Buick GMC
90 daysrisk line: 65
13-week cash · churn-weighted
amber = probability-weighted at-risk
Seller queue · price drops into buyable range
- 2019 Toyota Tacoma TRD$21,500 $20,050 47s
- 2020 Ford F-150 XLT$26,000 $23,500 1.3h
Everything the console does
eight working surfaces, one account record — no number appears twice with two values
Who to call today, and what to say
Eleven at-risk accounts ranked by expected loss × renewal urgency × save difficulty. Each opens vitals against benchmark, the exact drivers that moved the score, and a drafted outreach that cites live numbers — gated by a review checkpoint before send.
Open the run sheet →02 · Health ScoringA score that shows its math
Every score reconciles exactly: baseline 100 plus five signed drivers. The formula is published verbatim — weights, benchmarks, curve — and editable from a settings gear, re-scoring the whole book live as you drag.
See the formula →03 · SellersA live queue that answers first
Private sellers whose ask just dropped into buyable range, routed to the right rooftop with per-lead contact speed. Duplicates suppressed with receipts, and a sub-minute autonomous pilot running on three rooftops.
Watch the queue →04 · MarketsWhat a territory really yields
Modeled ceiling vs actual per market with competitor-overlap evidence the console computes on itself — including the scatter that keeps the analysis honest. The re-forecast conversation, armed.
Open the yield model →05 · OriginationFix the number promised at signature
Quoted-at-signing vs the defensible band vs actual, per market and by signing cohort, with an over-promise index. Most churn stories start at the quote — this is the guardrail that ends them.
See quote discipline →06 · CFOChurn priced straight into cash
A 13-week rolling forecast where at-risk renewals carry probability-weighted amber, with scenario toggles that drop real renewal invoices out of their landing weeks. Retention and finance as one program.
Open finance command →07 · Live AIClaude in the loop, never off the record
Regenerate outreach drafts in the rep's voice, stream a GM-ready weekly ROI report, or ask the console assistant anything about the book — every answer grounded in a live data snapshot, numbers narrated, never invented.
Try the AI →08 · BoardBuilt for leaving on screen
Auto-cycling executive slides — NRR trend, revenue saved, the win of the week, the biggest risk with a renewal countdown. Plus a view guide behind the ? button that explains every screen in one card each.
Open the board →Two divisions, one account record
The sales side works accounts; the finance side prices them. Both read the same record — health score → churn probability → cash — so CS and the CFO argue from one set of numbers.
Keep the cars coming — and the promise honest
For CS reps, sales leadership, and dealer GMs: triage the book daily, coach the right lever per cause, quote new markets at numbers the model defends, and win private sellers by answering first.
- Account health scorepublished formula · baseline 100 + drivers, exact
- 0–100
- Expected loss on at-risk bookARR × churn probability, per account
- $86.3k
- Speed to leadmedian first contact · sub-minute pilot share
- 3.1h · 4.4% <60s
- Offers per day vs benchmarkbook top quartile, trailing 90d
- 20/day
- Offer → purchase conversionbook median at modeled capacity
- 5%
- Units vs modeled capacitythe 50%-weight vital — yield truth, not effort
- per market
- Save rate · cost per save3 saves from 9 interventions this quarter
- 33% · ~$555
- Over-promise index at signaturequoted band ÷ defensible band, per market & cohort
- OPI
Cash that already knows about churn
For the CFO and controller: a rolling forecast where at-risk renewals carry probability-weighted amber, scenario toggles that drop real invoices out of their landing weeks, and a close that runs itself.
- Net revenue retention5.8 pts below breakeven — the number the program moves
- 94.2%
- 13-week contracted cashannual renewals landing in their actual weeks
- $1.8M
- Expected renewal lossfed live from Retention — same curve as the run sheet
- $86.3k
- Forecast accuracyup from 67% pre-model
- 94%
- Month-end closedown 2.9 days vs baseline
- 4.1 days
- Cost per invoicefrom $13.40 manual · AP runs as a secondary strip
- $3.10
- Revenue saved · early interventionsignal-to-outcome trail behind every save
- $63.4k
- Book bridgechurn concentrates below the $23.9k account average
- $3.5M ARR
A health score that shows its work
No black box: the exact formula is published in the product, every account's score decomposes to baseline 100 plus its drivers, and a settings gear lets you edit weights and benchmarks — the whole book re-scores live so you can pressure-test the model in the room.
Churn probability = (100 − score) ÷ 100 · expected loss = ARR × probability — the same curve feeds the run sheet and the CFO cash forecast.
score = round( 100 × Σₖ wₖ · rₖ^1.5 ) over k ∈ {units, conversion, offers, contact, seats}
r_units = min(1, units/mo ÷ modeled capacity) w = 0.50
r_conversion = min(1, conversion% ÷ 5%) w = 0.20
r_offers = min(1, offers/day ÷ 20) w = 0.10
r_contact = min(1, 120min ÷ median first contact) w = 0.10
r_seats = min(1, active seats ÷ licensed seats) w = 0.10Ten minutes in the console beats any slide.
Walk the five hero accounts, toggle the churn scenario, ask the assistant a hard question. Everything is live.