One batch. One flat price. One fixed number of dollars at risk. This document sets out every term of that pilot, including the parts that are usually left for a buyer's compliance team to find on its own.
This is a commercial term sheet, not legal advice, and not a binding agreement until both parties sign at section 15. We are not lawyers. Have your counsel read it, along with your own calling procedures, before the first dial.
The whole deal on one screen. Every line below is expanded later in this document, and nothing is buried.
| Pilot size | 250 records, drawn from a current batch of 1,483 wireless-verified call-ready candidates. |
| Selection | Matched to the buy-box the Buyer states in the live session and recorded at section 4. |
| Price | Flat, {{PRICE_PER_RECORD}} per record. Not usage-based. No per-call, per-minute, per-seat or per-outcome charge. |
| Total exposure | {{PILOT_TOTAL}}. That is the maximum the Buyer can spend under this term sheet. There is no second invoice. |
| Payment | {{PAYMENT_TERMS}} |
| Delivery | {{DELIVERY_DATE}}, as a single file, with per-record identifiers. |
| Working window | 14 days from delivery. |
| Exclusivity | One buyer per record. We never resell a delivered record to anyone. Section 6 states the limit of that promise plainly. |
| Replacement | Free, one for one, on any record that fails the stated buy-box on delivery. Definition at section 5. |
| Buyer obligations | Agreed contact-attempt discipline and per-record dispositions returned on a shared sheet. Section 7. |
| Dialing | Performed entirely by the Buyer, under the Buyer's own registry subscription, procedures and training. Section 12. |
| Renewal | None. Nothing auto-renews. Nothing rolls over. A second batch requires a new signature. |
| What both sides get | A measured benchmark of this supply against this desk, on this script, with these callers. That benchmark is the real deliverable of a pilot. |
The pilot is deliberately small. It is sized to be a decision one person can make in a room, not a procurement exercise.
The 250 records come out of a live batch of 1,483 call-ready candidates that already exists and has already been verified. They are not assembled after signature. The batch composition today:
Geography note, stated because it is the kind of thing a buyer finds later: 49 states and the District of Columbia appear in the batch. Vermont does not. Thirteen records carry no state value on file and are excluded from any state-filtered pull, or replaced under section 5 if one slips through.
Performance pricing sounds better and works worse at this stage. It asks a buyer to trust our integrity on a scoreboard we cannot audit, before we have shown a single funded outcome. A flat price on a small batch removes that argument completely. You pay a known number, you keep every dollar of the upside, and we earn the right to a bigger conversation with a result instead of a promise.
Every delivered record is a person-level contact record joined to a funding-intent signal. Field coverage is stated, not implied, and the guarantee at section 5 is written against these fields.
| Identity | First name, last name, age band. |
| Mobile | Ten-digit mobile number, line type confirmed wireless by live carrier lookup, carrier named on the record. |
| Personal email address. Present on 1,452 of the 1,483 records in the current batch. | |
| Location | Street address, city, state, ZIP. |
| Financial bands | Net worth band, household income band, homeowner flag. |
| Intent | Score category (High or Medium), matched topics across business loans, SBA, merchant cash advance, line of credit and equipment financing, and the window the signal was measured in. |
| Rank | Fundability rank, our internal ordering, with a plain-language reason string for why the record ranked where it did. |
| Identifier | A per-record id, used for dispositions at section 7 and replacement claims at section 5. |
Each record is a call-ready candidate. It is not a "lead", a "verified lead", a "warm lead" or a "hot lead", and we will not describe it that way in a room, in a file, or in an invoice. A candidate becomes something more only after your caller speaks to the person and confirms it. That distinction is the whole reason the price is flat and the batch is small.
The Buyer states its buy-box in the live session. It is written into this section before signature, and the replacement guarantee at section 5 is measured against exactly these values and nothing else.
| States accepted | {{BUY_BOX_STATES}} |
| Minimum net worth band | {{BUY_BOX_MIN_NET_WORTH}} |
| Minimum household income band | {{BUY_BOX_MIN_INCOME}} |
| Homeowner | {{BUY_BOX_HOMEOWNER}} |
| Other stated filters | {{BUY_BOX_OTHER}} |
| Always required | Mobile number present, line type confirmed wireless, record not previously delivered to the Buyer. |
If a stated filter cannot be met at 250 records from the current batch, we say so before signature and adjust the filter or the size with the Buyer, in the room. We do not quietly fill the shortfall with records that miss the box.
Any delivered record that fails the section 4 buy-box is replaced free, one for one, with no cap on the number of replacements.
| Wrong state | The record's state is outside the accepted list at section 4, or the record carries no state value at all. |
| Band miss | The record is missing the agreed net worth band or the agreed household income band, or the band on the record falls below the agreed minimum. |
| Not wireless | The record's number is not a wireless line at the time of delivery, confirmed by carrier lookup. |
| Duplicate | The record duplicates another record in the same delivery, or duplicates a record we previously delivered to the Buyer. |
This is a data-accuracy guarantee, not an outcome guarantee. No answer, voicemail, a bad mood, no interest, or a file that turns out not to be fundable are not buy-box failures. Those are the results we are both here to measure, and a supplier who offers to replace them is quietly telling you the price already includes the padding.
We cannot promise that no other company anywhere will ever contact the same person. The intent layer we source from is a subscription platform. Another subscriber could independently surface the same individual through their own pull, and no supplier who sources this way can honestly guarantee otherwise.
Any vendor promising universe exclusivity on behavioral intent data is either not sourcing it the way we do, or is describing something they cannot enforce. We would rather be the supplier whose exclusivity clause survives your counsel reading it word for word. What we promise is what we control, and we control it completely.
A pilot that is not instrumented teaches nobody anything. If a junior desk runs a stale script into 250 records for three days and stops, the contact rate will be low and neither of us will be able to say whether that was the data or the desk. The terms below exist so the result means something.
| Working window | 14 calendar days from delivery. |
| First touch | First attempt on every record within 2 business days of delivery. |
| Minimum attempts | At least 6 call attempts per record inside the window, spread across at least 3 separate days and at least 2 different dayparts, before a record is treated as unreachable. |
| One script | A single agreed script and a consistent caller group across the batch, so the benchmark measures the data rather than a change in approach halfway through. |
| Dispositions | A disposition returned for every one of the 250 records, on a shared sheet keyed by record id, updated at day 7 and finalised at day 14. |
Fixed values, so both sides count the same things: no answer, voicemail, bad number, wrong party, right party contacted, not interested, qualified, application submitted, funded, do not contact.
The shared sheet carries record ids and dispositions. It does not ask for the Buyer's pricing, underwriting criteria, borrower documents or anything else about the Buyer's business.
These four numbers are defined here and their targets are written in before delivery, not argued about afterwards. Set the targets from the Buyer's own historical numbers on its current data source. Ours are not the benchmark. Yours are.
| Metric | Definition | Agreed target |
|---|---|---|
| Contact rate | Records where a live human answered at least once, divided by records delivered. | {{TARGET_CONTACT_RATE}} |
| Right-party contact rate | Records where the person named on the record was reached and identified, divided by records delivered. | {{TARGET_RPC_RATE}} |
| Application submitted rate | Records that produced a submitted funding application, divided by records delivered. | {{TARGET_APP_RATE}} |
| Funded rate | Records that produced a funded deal within 45 days of delivery, divided by records delivered. | {{TARGET_FUNDED_RATE}} |
Whatever these four numbers come back as, the Buyer ends the pilot holding something it did not have before: a measured read on this supply, against this desk, on this script, with these callers, at a known cost per record. If the numbers are good, the next conversation is about volume. If the numbers are bad, we would both rather have learned it on 250 records than on 5,000, and we will say so in writing rather than explain it away.
The Buyer is taking first-mover risk on a supplier with no track record. That is real, and it deserves something in return that is not a discount.
| Buyer gives | If the pilot meets or beats the agreed right-party contact target at section 8, the Buyer provides one of the following, Buyer's choice: a 20 minute reference call with one prospective buyer we introduce, or written permission to use the Buyer's name and logo as a pilot customer. |
| Buyer keeps | Every funded deal, every dollar of margin on it, and the exclusive records themselves. |
| If the pilot misses | The Buyer owes no reference and no logo rights, and we confirm that in writing within 5 business days of the final dispositions. We do not ask twice. |
| Confidentiality | No performance figures, dispositions, funded amounts or internal numbers of the Buyer are published or repeated by us without separate written approval, whether or not the trigger is met. |
The question every buyer asks second is whether this can happen every week or whether the pilot drains the well. Measured numbers, taken from the live source, not projections:
At roughly 48,600 qualifying records published per day, a standing requirement of 5,000 records a month is comfortably inside the supply, with the constraint being our verification throughput rather than the pool.
| Field | Fill rate | Note |
|---|---|---|
| Personal email | 95% | Highest coverage field in the export. |
| Street address | 88% | Supports address verification on your side. |
| Any phone | 81% | Includes landline and non-mobile lines. |
| Mobile | 74% | Before our own live carrier verification. |
| Net worth band | 70% | Used for buy-box filtering at section 4. |
| Household income band | 68% | Used for buy-box filtering at section 4. |
| Homeowner flag | 65% | Used for buy-box filtering at section 4. |
The freshest intent available to us is published dated one calendar day back. This is same-day-published intent drawn from a single-day window. It is not real time, we do not describe it as real time, and any supplier telling you their behavioral intent is live to the minute is describing something else.
You will work out the ratio yourself, so here it is first. The 1,483 verified candidates came out of a pool of 100,737. That is about 1.5 percent. We throw away 98.5 percent of the window on purpose, and the discards are the point.
| Stage | Records | What happens here |
|---|---|---|
| Window pulled | 100,737 | Two-day window, US, High plus Medium intent, five funding topics. |
| Unique mobiles | 73,699 | Deduplicated to one record per mobile number. Everything without a usable mobile is dropped here. |
| Selected for verification | 1,500 | Ranked on fundability and field completeness, then the top slice is taken forward to live carrier verification. |
| Dropped at verification | 16 | Line type came back not wireless. Removed, not repriced or reclassified. |
| Call-ready candidates | 1,483 | Every one confirmed wireless by live carrier lookup. This is the batch your 250 come from. |
A high pass-through rate would be the warning sign, not this. It would mean the supplier is selling the raw window with a coat of paint and charging you to discover which 98 percent were unusable. The narrow funnel is where the price goes, and it is why the guarantee at section 5 can be objective instead of vague.
We supply data. You place calls. Those roles stay separate for regulatory reasons, not stylistic ones. This section states what the records are, what they are not, and which party carries which obligation, in plain terms, before your compliance team has to go looking.
Each record is a behavioral intent signal joined to a person-level contact record. The signal means the person's online behavior matched funding-related topics inside the measured window. It is an audience signal, and it is a good one.
The signal is not consent. It is not a form fill, an inquiry, a quote request, or a submission made to you or to us. It does not create an established business relationship. We hold no consent artifact of any kind and we do not represent that one exists. If anyone in this market tells you their intent data carries consent, ask them to produce the artifact.
We do not perform a federal Do Not Call registry scrub on these records, and we could not transfer one to you even if we did.
So the clean structure, and the only one we will sign, is that the Buyer scrubs under its own subscription. Records are delivered raw, for the Buyer's own screening.
| Party | Carries |
|---|---|
| Buyer | Subscribes to the registry in its own name. Scrubs the federal registry and any applicable state lists, plus its own internal do-not-call list, before dialing. Applies its own calling-hour rules, consent rules, recording rules and state licensing. Trains and supervises its own callers. Decides who gets called. |
| Supplier | Delivers records matching the stated buy-box. Verifies line type by live carrier lookup. Replaces buy-box failures. Places no calls, authorizes no calls, and operates no dialer. Does not resell a delivered record. |
We are not lawyers and nothing in this document is legal advice. The citations above are given so your counsel has the specific provisions in front of them. Have counsel review this term sheet and your own calling procedures before the first dial.
Stated up front, so nothing here is a surprise in diligence.
| We place no calls | No dialer, no agents, no call center, no live transfers, no appointment setting. Booked calls are not a product we sell, which is why nothing in this term sheet prices one. |
| We do not scrub the federal registry | And we cannot transfer a scrub to you. Section 12.3 explains why that is a legal constraint rather than a shortcut. |
| We hold no consent artifacts | Behavioral intent is an audience signal. It is not consent, not an inquiry, and not an established business relationship. |
| We have no track record yet | No case studies, no named reference buyer, no funded outcomes to point at. This pilot is the first one. The size of the batch, the flat price and the reference trade at section 9 all exist because of that, not in spite of it. |
| We do not price on your outcomes | Not in this pilot. We cannot audit your funded-deal reporting and would not ask you to open your books to us. See section 14. |
Nothing in this section is owed by either party now. Neither party is obliged to proceed. It is written down so both sides know what a good outcome opens up, and so nobody has to renegotiate from a blank page.
| Trigger | A funded rate at or above {{PHASE_TWO_TRIGGER}} on the 250 pilot records, measured from the final dispositions at section 7. |
| If met | The parties discuss, in good faith and with no obligation, a phase two on standing weekly volume. At that point a performance component becomes a reasonable thing to talk about, because there is a measured number on the table instead of a promise. |
| If not met | The pilot ends. Nothing renews. Both sides keep the benchmark, and the Buyer keeps every record it paid for. |
| Until then | This is a flat-price data purchase and nothing else. No gain share, no revenue share, no points, no equity, no standing commitment. |
By signing below, both parties agree to the commercial terms set out in sections 1 through 14 for this single capped pilot. This term sheet does not create any obligation beyond the pilot described here.
This document sets out commercial terms only. It is not legal advice, and we are not lawyers. Before signing, and before the first dial, have your own counsel review this term sheet, the allocation at section 12, and your calling procedures. If your counsel wants a term changed, send it back marked up. We would rather sign a document your lawyer has already read.