How it works
What lands on your desk, the whole loop.
- 01
Month-Zero Baseline, co-signed
Before we change anything, we co-sign a Baseline that records where your intake stands today, so every later number is measured against a line we both agreed to.
Safeguard: You approve the starting point; nobody moves the goalposts later.
- 02
Call arrives
Every intake call comes in through your existing phone system (CallRail, RingCentral, 8x8, Dialpad, or any system that records) or a manual upload.
Safeguard: Nothing changes about how your clients reach you, and nothing changes about the CMS you run it in: the desk sits on top of Filevine, CasePeer, Litify, or your own stack, with nothing to migrate.
- 03
Transcribed
Our transcription model produces a speaker-separated transcript.
Safeguard: The audio is deleted the moment the transcript exists.
- 04
Scored, and placed in a tier
Each call is scored against a frozen, calibrated rubric and placed in one of five plain-English signability tiers.
Safeguard: The rubric is version-locked, so a Tier 4 means the same thing every month.
- 05
Signable-case detection
A case is flagged when it reaches Tier 4 or above, it wasn’t converted, and it’s within 72 hours.
Safeguard: Pure, inspectable logic, not a black box.
- 06
Ali reviews every flagged call
A named analyst reviews every flag and the transcript moment behind it before it reaches your queue.
Safeguard: Not an autonomous model deciding on its own; a person who knows what a signable case sounds like.
- 07
Same-day Recoverable-Lead Alert
You get a one-screen Alert the same day. Lead value fades fast, so the play is a callback within the hour.
Safeguard: A specific, same-day SLA, because speed is the biggest lever on recovery.
- 08
Your staff make the callback
Every flagged case comes with the number, the evidence, and a suggested script; a person at your firm makes the call. During the beta that’s the whole play: no texts, no automation touching your callers.
Safeguard: We never contact your callers. A drafted, human-approved text assist is on the roadmap, gated on A2P 10DLC registration and seven compliance gates.
Human approval
A person on your team approves every message before it can send.
Opt-out
STOP / QUIT / REVOKE / OPT OUT / CANCEL / UNSUBSCRIBE / END (plus Spanish keywords) opt a number out instantly and permanently.
Global kill switch
One switch halts all sending across every firm.
Per-firm autonomy lock
A firm must be explicitly set to manual; no autonomous sending mode is enabled for any firm.
Per-firm kill switch
Each firm can halt its own sending independently.
Quiet hours
No messages 8:00pm to 8:00am in the recipient’s local time.
Test mode
Until A2P 10DLC is approved, sends are simulated and logged, never transmitted.
- 09
Monthly Missed-Revenue Statement
A two-page monthly Statement, signed by the analyst, shows what walked in dollars, the trend, and the Saved-Case Ledger. A standing 15-minute review call and a Right of Reply come with it.
Safeguard: Challenge any finding; your correction runs, unedited, in the next Statement’s Corrections column.
- 10
We grade ourselves against what you signed
Weeks later, we reconcile every call we graded against what your firm actually did with it: signed, declined, or referred. That tells us how often our SIGN was right, and our worst error, the cases we told you to pass that you signed anyway. We publish those numbers with their sample size.
Safeguard: The one number on your intake that isn’t self-graded is ours. See the Accuracy page.
- 11
Flat monthly fee
You pay one flat monthly fee tiered by call volume, the same whether you sign zero cases or fifty.
Safeguard: Never per case, per client, or per recovered dollar (Cal. B&P §§6151-6152).
For the intake department: this is on your side
We score the call and the process, not the individual’s job security. When the phones spike after a campaign, even a great intake team can’t catch every callback. That’s a staffing-and-timing problem, not a people problem, and the report says so. Scoring 100% of calls replaces the unfair 2% spot-check with an even standard for everyone.
This isn’t a gotcha, and it isn’t aimed at any one coordinator. At high volume, good cases slip; that’s math, not a character flaw. The desk gives your intake department proof of the workload, coaching clips built from your own team’s best calls, and a monthly scorecard, per office, that shows the improvement so the credit lands where it’s earned.
It’s proof of workload, the case an office manager usually has to make from memory: you’re not short on effort, you’re short on hours. The team sees its own numbers first, and a person on staff approves every send. It protects careers; it doesn’t threaten them.
Your first 30 days, by the clock
- Owner: about 20 minutes total, a kickoff call and sign-off.
- Office manager: about 2 hours of setup, mostly exporting or forwarding call recordings.
- Intake team: about 5 minutes a day confirming the triage queue, which fills itself from your recordings so no one re-types a form.
That’s the whole lift.
We measure speed, from your own recordings
We measure time-to-answer and time-to-callback on your own calls, because speed decides conversions: the faster a firm responds to a new inquiry (minutes, not hours) the more of those callers sign. We show you where your own response times are costing you signable cases.
What months 2-12 look like
The first audit finds the biggest leaks. After that, the desk becomes the one number on your intake that isn’t self-graded: an independent monthly scorecard your intake operation is measured against, across every office, the way you’d never drop your malpractice carrier. Every statement shows whether the leak is shrinking, your team gets credit for the improvement, and new leaks get caught as your marketing and staffing change.
- A statement that trends over time: Your monthly Recovered statement shows the leak shrinking, not just a static snapshot.
- An accuracy scorecard on us: The Our-accuracy page grades our own calls against what your firm actually signed, so the one number on your intake that isn’t self-graded is ours.
- Coaching clips from real calls: Short training clips built from your team’s own best calls.
- New-leak detection: As your marketing and staffing change, new leaks get caught.
- Save-protocol conversion tracking: Once A2P 10DLC clears, we track how many saved cases the protocol actually recovers.
Some of the above is live today and some is rolling out with the founding cohort. We label which is which in your kickoff, and we never bill for something that isn’t running yet.
Spanish-language calls, handled honestly
We detect the call’s language automatically. Today our scoring rubric is calibrated and validated on English-language calls; we won’t claim validated Spanish-language scoring until we can show the test corpus behind it, so during the beta Spanish calls are reviewed personally rather than scored automatically. It still matters in California: 48.4% of Los Angeles County residents are Hispanic or Latino (U.S. Census Bureau, 2020 Census redistricting data), Spanish-first callers are where English-only intake leaks most, and a signable Spanish-speaking caller is worth exactly as much as any other. Ali is bilingual and reviews flagged Spanish calls himself.
See it on your own calls.
Upload up to 10 recent intake calls and get a dollar-quantified report of the signable cases that didn’t sign, with the evidence behind every flag.
Get your free Leak Audit