Stop selling the machine. Point it at a market that already buys conversations every week.
Capital advisory pays big and pays slowly. That is the whole reason the pipeline feels stuck: a $40K engagement needs ninety first meetings, and a raise takes quarters. The fix is not a better version of that lane. It is running the same engine into a market with a weekly budget line for exactly what it produces, and selling that output by the unit.
The ask, in your words
Kevin · 2026-09-07
KevinThinking of investing into our own engine and start closing deals for 1 or 2 different industries. Fed up with those slow and low volume pipelines. Want to find a way to crush it.
KevinAdditionally there is still the service business which we could start closing deals for the service itself, but I think too low profit margins.
KevinI want to crush it but feel super stuck right now with low volume of leads and clear definition of product.
LukaAgreed on all three. The service business sells the build at cost. A desk sells the output at margin, every week, to buyers who already pay for it. Here is where to point it, ranked on evidence we actually have.
The product, in one sentence
We sell qualified conversations with people who are looking for what you sell this week, delivered daily, priced per unit, and guaranteed by count.
Not a build. Not software. Not a monthly retainer for effort. A buyer picks a unit (a dial-ready record, a live transfer, a booked meeting), sets the gates, and pays only for units that clear them. One engine and one database behind every desk, run by us. The buyer gets a dashboard and a calendar. That is the product, and it is the same sentence in every industry below.
What you say to a buyer "We put verified, in-gate prospects on your calendar or in your dialer inside fourteen days, or we keep working free until we do."
Nine candidates, three axes, scored on what we can prove
Industry
Ticket and margin
Days to first check
Supply we can prove today
Verdict
Business funding (MCA, working capital)Funders, ISOs and brokers buy leads and transfers weekly and decide in one call. Engine already built and running daily.
4
5
5
START HERE measured190,100 businesses/day searching, 2,300/day decision makers in band.
Individual health insurance (ACA + Medicare)Medicare AEP opens Oct 15, ACA open enrollment Nov 1. Agencies stock up on leads in September and October.
5
3
3
SECOND, on a clock windowPricing and script already set. Consumer supply count still to measure.
Consumer credit repair + debtPhone floors of 20 to 30 reps, close 6 to 15%, Meta is their only source. They buy the in-market file.
3
4
4
THIRD, file desk measured42,501/day clear a real buyer's full buy-box. Response rate on that file never tested.
M&A sell-side (business brokers)Market pays $250 per qualified seller meeting and six figures per closed fee. Proposal live with one broker.
5
2
3
UPSIDE LANE, not a volume lane10% at signed plus 10% at earned. Months to a fee.
Solar, roofing, home servicesBooked solar appointments sell $200 to $400. Roofing appointments $175 to $200.
4
3
1
untested by usRich market, zero supply measured on our side. Not this month.
Tax relief, debt settlementAdjacent to the credit audience, live transfers price well.
4
3
2
untested by usFold into the credit desk once round one is dispositioned.
Mortgage, refinanceRate-driven demand, heavy compliance, crowded.
3
3
1
untested by usPark it.
Personal injury on cold intentRan it for real. The intent file was legal-content readers, about 1% real accidents. 4,332 cold texts, one yes, zero signed.
5
1
0
NO verdict 2026-09-02Only works as Meta geo ads plus a human intake. Different business.
Capital advisory as a volume laneYour current pain. 50 booked, 25 to 30 show, 5 to 10 qualified, one $40K fee per ninety meetings.
5
1
2
KEEP for upside onlyKeep the desk. Stop expecting it to fix cash flow.
Scores are 1 to 5. Measured means we counted it on a real export or a real run. Window means the timing is public and fixed. Untested means the market data is sourced but we have never pulled a record for it. The three rows we pick are the three where a check can clear inside thirty days on supply we already hold.
One engine, every desk. Measured on the funding lane, one calendar day, US
Everything left of the green box is one calendar day of supply on the funding lane, counted on a real export. What falls out at each step is shown in red because the credibility is in what we discard, not in what survives. The three units on the bottom row are the only things a buyer ever pays for, in any industry. Swap the topics and the gates and the same picture holds for insurance and credit.
Desk one. Start here this month
Recommended first desk
The Funding Desk
Fastest check in the building. The engine runs daily already, the funnel is measured, the buyer decides in one call and pays weekly, and the market already prices every unit we sell. Your syndicate and your capital network are full of people who know an MCA funder or a broker.
Who writes the checkMCA funders, ISOs and brokers. They buy every week, they already spend on leads, and they know what a $100 transfer is worth.
Supply, measured190,100 businesses searched for money in one day. 2,300 were decision makers at companies in band. ~1,000 clear a real buyer's two gates.
Quality, measuredDelivered sample: median $1.0M/month revenue, median 26 years in business, smallest merchant $250K/month against a $100K floor. ~80% carrier-verified mobile.
Unit
Our price
What the market pays today
Who does the dialing
Wedge: the proof run, 2,000 records
$4,000, $2 each
Aged MCA $0.05 to $0.50, UCC $1 to $5, real-time $50 to $200
The buyer. Four days to delivery, any record that fails a gate replaced free.
Dial-ready record after the run
$2.50 to $5
Real-time leads $50 to $200 from list vendors
The buyer. Rate set by their own dispositions.
Live transfer, gates confirmed on the call
$125 in a package proposed
$75 to $150 per transfer, $30 on Fiverr with no gates
Our desk, human or AI with disclosure, then handed to their closer.
The Records Desk, monthly
$7,500 for 3,000 records a month proposed
$2.50 a record, same-day signal, one buyer per record
The buyer. Any record that fails a stated gate is replaced free.
The Funding Desk, monthly
$15,000 for 120 transfers proposed
Starter $6,500 for 50, Full $25,000 for 220
Floor from month two. Miss it, we dial the gap free. Transfer price is set inside the pilot, never before a buyer has priced a real one.
DAY 1 to 3Buy-box on a call: states, revenue floor, time in business, industries they will not touch. $4,000 in.
DAY 4All 2,000 records land on their desk with DNC and litigator status on every row.
DAY 5 to 12They dial. Dispositions come back. We recut the file against what closed.
DAY 13 to 14First live transfers from the recut file. Rate for month one set from their numbers.
DAY 15Desk invoice one. Second buyer opened from the syndicate while the first one is still hot.
What kills it. The market is commoditized at the bottom, so we never compete with $30 transfers. We sell the same-day signal, the gates, and one buyer per record. Second: DNC. In a 58-record sample, 45 were listed. We put the status on every row and the buyer's counsel sets the policy, we never scrub for them. Third: transfer volume is dial-capacity bound, so we sell records first and grow transfers as DIDs and desk hours grow, never the other way round.
Desk two. On a clock
Window opens Oct 15
The Health Desk (ACA + Medicare)
Medicare annual enrollment runs Oct 15 to Dec 7. ACA open enrollment runs Nov 1 to Jan 15. Agencies buy in September and October, and in-season leads price 20 to 30% above the rest of the year. Every week we wait is a week of the most expensive lead window in the industry that we miss.
Who writes the checkIndependent health agencies with 5 to 20 agents, and networks of them. One network we already know runs 150 to 175 agencies at 50 to 125 calls a day each.
The buyer profileAges 26 to 64, income $45K+, 1099, gig, small business, no employer plan. Triggers: turned 26, lost a job, started a business. The SMB owner is the multiplier: owner, then family, then staff.
SupplyFree state filings already produce 1,251 new business owners a day from one state. Consumer intent topics for health not yet counted. pending, day 5
Unit
Our price
What the market pays today
Who does the dialing
Trial pack, 25 leads
$50 each, $1,250
ACA leads $32 to $85
The agency. This is the falsifier: if 25 do not convert, no one buys 100.
Weekly desk, the floor
$3,000 a week for 40 leads
Agencies already prepay weekly for click-to-calls at $150 with no name attached
The agency. Weekly prepay, no contract, same-day replacement.
Qualified lead, weekly prepay
$75, then $100 to $125 once close rate is proven
Health live transfers $40 to $90 in season
The agency. Same-day replacement on any bad record.
Live transfer add-on
+$25 to $50
Medicare interest-verified transfers $110
Our desk, human dialers only. Never quotes, never advises, transfers at peak interest, state-matched.
SMB-owner lead
$150 to $200
Pre-vetted Medicare tiers $250 to $300
Our desk, human dialers. Owner, family and staff on one call.
DAY 1Register a rotating pool of numbers with the carriers, about four business days, and staff two human dialers. Both go first.
DAY 2 to 5Build the consumer audience, count it, set the age cut. Sell the first 25-pack to one agency on the count.
DAY 6 to 7120 to 150 validation calls. Two numbers come out: dial to contact, contact to yes. They set the price.
DAY 8 to 14First weekly prepay. Second and third agency opened on the first one's numbers.
OCT 15Live inside the Medicare window with three agencies on weekly prepay. That is the goal date.
What kills it. AI voice to a consumer is an artificial-voice call under TCPA and needs written consent in 47 states. So the outbound side of this desk is human dialers from day one, not as a fallback for later. Brooke does not touch a consumer dial here until a written-consent flow exists and has been tested. The cleanest volume is consented inbound from lead-form ads, which Brooke can work. Second: carrier spam labels. The fix is registered numbers, at most 15 dials a day per number, area-code matched, business hours. We solved both already for the last insurance deal. The only way to lose the window is to start in October.
Desk three. The file desk
Buyer already said yes
The Credit Desk (credit repair + debt)
A credit repair shop with 26 phone reps, adding ten a fortnight, $99 a month product, closing 6 to 15% of leads, spending $25K a month on Meta as its only source, told us on a call: "we find a way to make that work, we can make the first purchase, no question." This is a file business, not a dialing business, so it needs no desk hours from us.
Who writes the checkCredit repair and debt settlement shops with a phone floor. Their humans dial. Their problem is a second source that does not poison the first one.
Supply, measured340,542 raw in 48 hours across five credit topics. 85,002 clear the full buy-box: mobile, email, income known and over $45K, under 65. That is 42,501 a day.
The scarce thingOnly 2.3% is top tier, about 978 a day, roughly one buyer's weekly capacity. Sell that as the premium, never the bulk.
Unit
Our price
What the market pays today
Who does the dialing
Round of 1,000 buy-box records
$4 each, proposed
Their Meta cost per lead is $6 to $15. Debt relief leads run pennies to $75.
The buyer. Six rounds, alternating who funds them, dispositions between every round.
Top-tier record
$8 each, proposed
Scarce. 978 a day exist.
The buyer.
Response-validated lead (replied yes by text)
$15 to $25, proposed
Still under their $6 to $15 Meta cost once close rate is counted, and it arrives already talking
Our sending stack qualifies, their reps close. This is the premium unit, sold after round one proves the signal.
The sending stack (the deal closer)
$1,500 a month, proposed
Nobody sells this next to the file.
Separate sending identity and number pool, so a STOP storm on bought data never throttles their Meta lane. Opt-outs suppress globally.
DAY 1Round one funded by the buyer, $4,000. Audience cut to their buy-box, 65+ excluded.
DAY 3Sending stack approved and live, separate from their CRM numbers.
DAY 5Dispositions on round one. First time the signal is response-tested. Truth arrives here. No monthly floor is sold before this clears.
DAY 6 to 25Rounds two to six, alternating funding, file recut each time against what closed.
DAY 30Ongoing rate set from their logged numbers. Second shop opened on the first one's close rate.
What kills it. The signal has never been response-tested. Nobody has dialed or texted this file yet. That is why round one is small and why the rate is set after, not before. Second: line-type and DNC checks are not yet run on it and must be before a single message goes out. Third: this is a lower-ticket desk with a buyer who already knows what a lead costs, so it is a margin business on volume, not a premium desk. Good third leg, wrong first leg.
The upside lane, kept separate on purpose
Second act
Sell-side M&A and the advisory desk you already run
This is where the six-figure fees live, and it is also exactly the lane that made you feel stuck. Keep it, price it on the upside, and stop asking it to pay this month's bills. The three desks above fund the wait.
The model already shipped$7,000 build, then $7,000 / $12,000 / $20,000 a month by intensity, and the only reward: 10% of the engagement fee at signed plus 10% of the transaction fee at earned. Both payable after the money is in their account.
Market referenceBrokers already pay $250 all-in per qualified seller meeting. We do not compete on that. We are the only vendor in their inbox that loses if a mandate does not sign.
Why it is not firstThirty days to the first meeting, months to a fee. It compounds beautifully and it cannot fix a cash problem. Two of these running while the volume desks pay the rent is the right shape.
What each side puts in, for a vertical desk we own together
Kevin
The run cost of the desk: feeds, enrichment, registered numbers, dial minutes and messaging. Roughly $5,000 to $8,000 a month per desk at 1,000 to 5,000 finished leads. cost sheet Aug 30
The buyer conversations. Three funders, three agencies, three credit shops, from your network and the syndicate. You close, we deliver.
A closing calendar that is actually open, and a human on it, so a transfer never lands on voicemail.
Cash and buyers
Luka
The engine, already built and running daily. Adapting it to a new lane is an ICP file and a script, not a rebuild.
About 100 hours of unbilled build per desk, priced at $100 an hour so it disarms rather than inflates: $10,000 of work nobody invoices.
Brooke, the dashboards, the recordings, the transcripts, the ops. Every desk is dedicated, nothing shared with any other client.
Build and ops
Revenue splits 50/50 on desks we own together. Neither of us is paid for effort. You risk cash, I risk the build, and we are both made whole at the same moment: a buyer pays for a week of units and comes back for the next one.
What one desk looks like at three buyers
Gross, funding desk
$45,000
three funders on The Funding Desk at $15,000 for 120 transfers each. assumed
Run cost
$8,000
top of the cost-sheet range at volume, feeds are the biggest line. cost sheet
Left over
$37,000
per month, per desk, before the split. That is a month-one number, not a year-three number.
Versus the service business
$7,000
one time, at roughly our cost, and then it is their engine. That is the low margin you felt.
the service sells the build once. the desk sells the output every week.
The one mistake to not make
Biggest bet against this plan
Opening three desks at once and proving none of them. One desk, one buyer, one invoice, then the second buyer.
Your own message is the tell: stuck on low volume and no clear product. Three offers at once recreates that exact problem inside the new desk. So: funding only for the first 30 to 45 days, to first invoice and a repeatable playbook, then clone the same engine into health timed for Oct 15, then credit. The only thing that starts early is the health desk's number registration and dialer hiring, because the carriers' four-day clock and the enrollment window do not wait for us.
Next step: you name one funder or broker you can get on a call this week. I have the buy-box questions, the console and the proof-run terms ready. First records land four days after the call.