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 or a manual upload.
Safeguard: Nothing changes about how your clients reach you.
- 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
A human approves the follow-up
Your staff reviews, edits, or rejects every ≤320-character follow-up (with “Reply STOP”) before anything sends, and each message clears seven compliance gates in order.
Safeguard: No autonomous sends, ever, in pilot. The first failed gate stops the message.
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; there is no autonomous mode.
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
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 team: 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. High-volume intake means good cases slip. That's math, not a character flaw. The desk gives your manager proof of the workload, coaching clips built from your team's own best calls, and a monthly scorecard 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 on the triage queue.
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 your standing intake QA function: every month your statement shows whether the leak is shrinking, your intake 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 missed-revenue statement shows the leak shrinking, not just a static snapshot.
- Intake-team scorecards: Improvement trending that makes the manager look good, so credit lands where it's earned.
- 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, analyzed natively
We analyze Spanish-language intake calls natively, not translated afterward. In the Los Angeles metro, 34.5% of people age 5 and older speak Spanish at home (USAFacts, 2019-2023 American Community Survey), and a signable Spanish-speaking caller is worth exactly as much as any other. Ali is bilingual and built the scoring against real calls in both languages.
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.
Run your free Intake Quality Audit