This hub offers practical guides and best practices for advisors looking to successfully introduce companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia (outside the U.S., Canada and the EU, at least 70% of the data must be in English) to SourceX, a direct licensing partner for approved business data for AI training and evaluation. Here, you'll discover step-by-step instructions and strategic insights for introducing companies to SourceX, which licenses data directly from companies and manages the agreement, rights review, data preparation, delivery and payout, subject to acceptance at Closing and agreed terms. These resources are designed to help you navigate the process efficiently, focusing on companies with 50+ full-time employees at peak. Unlock the potential of valuable data partnerships by leveraging these comprehensive guides.
Practical guides to successful company data introductions
Short, task-focused guides on identifying suitable companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia (outside the U.S., Canada and the EU, at least 70% of the data must be in English), recognizing valuable proprietary data, and making an introduction the decision maker will welcome.
- Step 1Share your linkSend your personal link to a company you know.
- Step 2Company appliesThe company applies itself at /apply.
- Step 3Company signs and is paidThe company signs, delivers its data and is paid by SourceX under its agreement.
- Step 4You get your rewardthe greater of $50,000 plus 5% and 15% of the actual supplier payout, with no cap. A reward becomes payable after SourceX has paid the referred company under its signed agreement and the payout conditions in the program terms are met, including finance approval and your tax and payment details.
Common questions
What kind of companies should I introduce to SourceX?
You should focus on introducing companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia (outside the U.S., Canada and the EU, at least 70% of the data must be in English) that have 50 or more full-time employees at peak. These are the companies that can unlock the potential for valuable data partnerships with AI labs and data buyers.
What is the purpose of these guides?
These guides offer practical instructions and strategic insights for advisors to successfully introduce companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia to SourceX. They help navigate the process efficiently and identify valuable proprietary data for AI labs and data buyers.
Who benefits from SourceX data introductions?
AI labs and data buyers benefit from approved business data that SourceX licenses directly from companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia for AI training and evaluation. Advisors can also benefit by introducing companies to SourceX as their direct licensing partner.
What kind of information do these guides provide?
These guides provide step-by-step instructions and best practices for identifying suitable companies in the U.S., Canada, the EU, the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia. They help in recognizing valuable proprietary data and making introductions that decision makers will welcome, facilitating data transactions for AI.
Related pages
- How to model a one-time receipt, like a data license, in a 13-week cash flow forecast
- 2021-vintage private equity funds: what to do with companies bought at peak multiples
- 3PL and logistics M&A in 2026: what buyers want and what sellers should preserve
- 3PL valuation multiples in 2026: what buyers pay for in a logistics business
- A/E industry M&A trends in 2026, and what selling firms should prepare
- Acquiring a business from a retiring owner: assessing decades of records before they go
- The investor dinner conversation that leads to a useful introduction: six permission-first cards that let a guest raise company records, hear who could be paid and drop the subject, with nothing moving until the CEO writes
An investor dinner can surface a useful data introduction, but it cannot make one. Let the topic come up only when a guest raises it, say once who could be paid, send every question about a company to that company's CEO, and stop there. The introduction begins when the CEO writes to you after the dinner and asks for it. The six permission-first conversation cards below cover that path, from the first question to the morning-after email, and keep pitches, lists and promises off the table.
Read → - The Jay Alix Protocol and CRO retention: where does a referral reward fit?
The Jay Alix Protocol is the U.S. Trustee Program's approach to retaining chief restructuring officers, and it centers on disclosing connections and compensation. A SourceX referral reward connected to the debtor's data should be raised with the CRO's counsel for disclosure, and counsel decides whether anything more is needed.
Read → - The joint office hour for two venture portfolios: a 21-day runbook in which CEOs bring the questions, every company has one follow-up fund and nothing reaches SourceX until a CEO asks
A joint office hour lets the CEOs of two venture portfolios ask how licensing company records for AI training works, in one session, without turning attendance into a referral. Settle three things before any invitation goes out: who chairs, whether SourceX staff attend, and which single fund follows up with each company, including companies both funds back. Nobody brings or collects records, no attendee list goes to SourceX, and an introduction is submitted only for a CEO who asks for one in writing afterward.
Read → - The M&A advisor's handoff to the current operating team: a three-part briefing packet that gives the CEO a decision, gives system owners the questions SourceX will ask, and leaves the deal room behind
Hand the decision to the CEO who runs the company today, and hand the groundwork to the operating team who would answer SourceX's questions, in one packet that carries only facts the company approved. Before closing, that CEO is the target's; after closing, it is the acquirer's. The packet names the decision waiting, lists the descriptive questions system owners will be asked, and states what the advisor keeps: data-room material, price and buyer correspondence never cross. Then the advisor steps out on a date written into the packet.
Read → - The member benefit case for a data partnership: a benefit-and-burden scorecard that rates each member segment from the member's side, so the board offers it only where members gain more than they carry
A data partnership is a member benefit only for members who can use it, and only where it costs them less than it gives them. Score each member segment from its own side: the benefit (a neutral explanation, a private fit check, a route its CEO controls) against the burden (time, legal questions it must pay to answer, the chance of reading your mention as an endorsement, the letdown of not fitting). Offer it where the benefit clearly wins, run a small opt-in pilot with a way to say no, and treat declines and complaints as evidence.
Read → - The member feedback loop that improves referral fit: a four-step cycle that codes each optional reply about the notice, changes one rule the association owns per round and records nothing about any member's company
Ask the members who took part one optional question about the association's notice, never about their company. Sort every reply into a fixed code, tie each code to a lever the association controls, such as the notice's first line, who receives it or how often it is sent, and change one lever per round. SourceX's criteria, its review and each CEO's decision stay outside the loop, and nothing about a member's records is collected, for SourceX or for the association.
Read → - The member introduction that preserves an existing advisor relationship: an advisor-aware email that asks the CEO who has already raised SourceX before any link goes out, and lets the CEO pick the route
Ask before you introduce. When a member's CEO asks the association for an introduction to SourceX, reply first with one question: has an accountant, consultant, IT provider, lawyer or other advisor already raised SourceX with you, or put you in touch? Then let the CEO pick the route: through that advisor, through the association, or straight to SourceX. Copy the advisor only if the CEO asks. One introduction can be credited, an earlier one may already exist, and the advisor relationship belongs to the member.
Read → - The message review that removes false urgency: find every time claim in a referral draft, name whose clock it runs on, then date it or cut it
False urgency is any time pressure in a referral message that the recipient's company did not create. Before a draft goes out, mark every time phrase, ask whose clock it runs on, and keep it only if it names a dated, documented event at that company or one of its vendors, such as a system retirement or a retention purge, in words the CEO can check. Market hype, invented deadlines and the partner's own 90-day attribution window never belong in the message.
Read → - The minimum information needed for a useful referral: a nine-field intake specification that lets the company describe itself, sends SourceX only what an introduction record must name, and keeps what a fund learned in confidence out of every field
A useful referral carries far less company information than most fund intake forms collect. SourceX's sample partner agreement says each introduction record "must identify the company, domain, Partner, opportunity, introduction method and timestamps," and the company's own application asks for everything else, from what it does to which systems it runs. A fund's intake can stop at nine fields: six that leave the fund once the CEO agrees, and three that stay inside for approval and conflicts. Board packs and information-rights reports fill none of them.
Read → - The MSP referral pilot that protects client trust: a 120-day trust-and-fit measurement plan with tripwires written before the first ask, a holdout group and a readout that counts the clients who said no
Run the pilot as a test of whether you can raise data licensing without denting client trust, not as a forecast of earnings. Choose 8 to 12 accounts with a dated system event and a reachable CEO, write the trust tripwires, fit measures and decision rule before the first ask, keep a holdout group you do not ask, and read out at day 120, when every introduction made in the first 30 days has had its full 90-day attribution window. Revenue stays out of the plan: SourceX publishes no conversion rates, and rewards arrive later.
Read → - The network concentration report every channel strategy needs: a stage-by-stage study design that measures how few relationship owners carry each step from a named company to a credited application, keeps every failed ask in the count and has no results yet
A referral network concentration report should show, stage by stage, how much of a network's progress rests on a few relationship owners: companies named, CEOs asked, CEOs who agreed, verified applications and credited companies. Measure each stage with the top-three share, the effective number of owners and a remove-one test, keep every declined and unanswered ask in the count, and pool networks only with written consent. This page is the protocol. No network has been measured, so no results exist.
Read → - The network health dashboard beyond registered partner count: four signals borrowed from site reliability engineering, and a quality-adjusted partner count for the headline
Registered partner count measures sign-ups, not health. Put four signals on the dashboard instead, adapted from Google's site reliability practice: traffic (unique companies proposed), errors (duplicates, fit failures and introductions the CEO did not choose), latency (days from proposal to a screen decision, and from the CEO's yes to a verified application) and saturation (screening hours used against hours available). Headline one number: partners whose last 90 days produced a unique, fit-passing, non-duplicate proposal.
Read → - The network review that finds owners instead of generic inboxes: a six-step workflow from the addresses your firm holds to the colleague who knows each CEO
Most addresses a service firm holds for a client reach a function, not the person who decides: billing, IT, support or a contact form. An owner network review maps each account from the best address you hold to the owner and CEO on public record, then to the colleague in your firm who actually knows that CEO, and asks that colleague first. Accounts reachable only through a role inbox are parked, never emailed.
Read → - The networks that learn fastest may find better data opportunities: a cohort-by-cohort test that separates learning from filtering
A network has a learning advantage only if its later introductions fit better for reasons it can name: sharper first questions, better-calibrated fit predictions, owners who understand the request before the first call. Volume is not learning. Write a fit prediction before each introduction, compare it with the stage SourceX records, debrief without company details and change one question between cohorts. If later cohorts predict better on a similar mix of companies, the network is learning. If fit rates rise only because it introduces fewer, easier companies, it is filtering.
Read → - The old company brand and the current rights holder: a brand-to-rights-holder crosswalk that follows a familiar name through renames, sales, wind-downs and revivals to whoever controls the records today, and stops when nobody can be named
Separate three trails before anyone is contacted: who owns the old name now, which legal entity's employees created the records, and who controls that entity or its estate today. Only the third decides. A buyer of the brand alone has no say over records it never acquired. After a stock sale, the acquirer's CEO decides. For a closed or insolvent company the CEO may not be enough, and whoever controls the estate decides. When the trails cannot be joined to a named controller, the file stops.
Read → - The one-page data referral brief a founder can approve: fourteen lines, each tagged with who said it and whether it is a fact, an estimate or an open question, with no records attached
A data referral brief is one page that the company's CEO approves line by line before anything about the company reaches SourceX. It holds metadata only: full-time headcount at peak, record types, years, systems, who owns the records, known constraints, what the CEO wants to learn and the questions nobody can answer yet. Every line is tagged with who supplied it and whether it is a fact, an estimate or an open question. Nothing from board materials, and no records, samples or logins, goes on it.
Read → - The operating handoff when a partner firm merges: a continuity plan sorted by who signed the partner agreement, which keeps each firm's introduction records as written, sends SourceX one acknowledged notice and leaves what the documents do not cover to SourceX
A merger does not rewrite a referral. Each introduction record names the Partner it was made under and carries its original date, and the 90-day window keeps running from that date. The Channel Partner Agreement itself can pass to a successor without SourceX's consent only on notice in a merger, acquisition or sale of substantially all relevant assets, and only where the successor assumes its obligations; any other transfer needs SourceX's written consent. Keep both firms' records exactly as written, notify SourceX in writing, and leave to SourceX whatever the documents do not settle.
Read → - The operational history behind a fast-growing support team: an eight-field worksheet that dates the tickets, playbooks and QA trail before anyone calls it a data opportunity
A support team that doubled tells you tickets are piling up; it does not tell you how many years of resolved work the company holds, in which system, or who owns it. Treat the hiring as a prompt to date the support history: when the helpdesk started, which migrations broke the trail, whether playbooks and QA reviews exist, and whether employees or an outsourcer do the work. Introduce the company only when that history, the rights and a willing CEO line up.
Read → - The opportunity cost of another founder program: a tradeoff scorecard that prices platform hours before any referral income
Another founder program is paid for in platform hours taken from the programs founders already use, so price it in those hours before anyone mentions income. Score it against the program it would displace on founder pull, usefulness, hours per founder helped, reach in your portfolio, relationship risk and reversibility. A data licensing referral track deserves a slot only if it beats the weakest current program with income set to zero. Referral rewards are uncertain, arrive late and should never decide the vote.
Read → - The overlooked dataset in a customer-success handover: unresolved promises
Commitments made at a customer-success handover link to outcomes through four records: the handover note that states the promise, the success-plan objective or task that tracks it, the support cases and escalations raised when it slipped, and the renewal or churn record at term end. The records matter as AI training and evaluation data only when a shared account ID, and ideally a commitment reference, joins them. Unresolved promises are the most telling part of that chain.
Read → - The overlooked reasons good data introductions fail: a complete-funnel map that records where each one stopped and the earliest point its cause could have been known
Most good introductions that fail stop for a reason that could have been known earlier than the stage where it surfaced: the person who said yes was not the CEO, the records belonged to clients, a colleague had already emailed SourceX, the application came after the 90-day window, or the founder heard about the partner's reward from someone else. Log every introduction, not only the memorable ones, with two stages each: where it stopped and where its cause first became knowable. The gap between them is what you can fix.
Read → - The owner's route from 'this might be valuable' to a qualified conversation: six stops, a real exit at each, and how to choose the first call
An owner is ready for a qualified first conversation when three things are true: they can say why they are exploring, they can describe which records the company itself created and who owns them today, and the CEO wants the discussion. Everything else they are unsure of goes on a written question list, and no files leave the company. As the network owner, walk them through six stops in order, offer a real exit at each one, and leave technical licensing decisions to SourceX's own assessment.
Read → - The partner advisory council that improves the program: a one-page charter that keeps named companies out of the room and answers every recommendation in writing
A referral partner advisory council improves the program when it advises on what the program owner can actually change, such as criteria wording, intake forms, support routing and training, and stays out of what it cannot settle: rewards, attribution, and whether a named company fits. Give it a short written charter, seats that include your weakest cohort, a rule that no company is named in the room, and a loop in which every recommendation gets a dated written answer and is checked one quarter later against a number you named in advance.
Read → - The partner brand guide for direct licensing language: an approved-language glossary proposal that rests every word about SourceX and the deal on a sentence SourceX published, retires the words that misdescribe it, and marks what only SourceX can approve
Describe SourceX in the words its own pages use. A company licenses its records to SourceX, which is its counterparty from agreement through payment and licenses the de-identified data onward to AI developers in its own name. The company keeps ownership of its originals, the license is exclusive for AI training, and SourceX says no de-identification method is perfect. Retire "sell", "marketplace", "anonymous", "pre-approved" and any AI company's name. Treat your glossary as a proposal: every entry cites a published sentence, and copy that names SourceX publicly waits for its written authorization.
Read → - The partner check-in that produces better introductions: a 30-minute, six-question agenda that starts with outcomes, catches overlap before anyone asks and ends in a decision
A partner check-in produces better introductions when it runs on questions about the partner's network and recent outcomes rather than on a pitch. Ask, in a fixed order: what happened to their last introductions and why, which CEOs they could ask next and which fit facts they already know, whether anyone else in your network knows the same companies, how they describe the reward, what got in their way, and whether to continue. Keep it to 30 minutes, write down one action for each side, and set how often you meet by the time you actually have.
Read → - The partner communication plan after program terms change: a versioned update rollout checklist that starts from the terms' own change log, finds everyone who heard the old wording and tells each of them only what changed, in a fixed order
A new terms version is a communication task only where someone relied on your old words. Start with section 11 of SourceX's terms, which says what each version changed. List every statement your team made that rests on the changed section, find who received it, and send each audience a short dated note in a fixed order: people about to act on your words first, then your own introducers, then published material, then the log. Quote the new wording and link it. Never tell anyone how the change applies to their own signed documents.
Read → - The partner enablement test before a large network rollout: a knowledge-and-fit pilot in six chapters, allocated by chapter, that audits what members actually wrote to CEOs, follows their introductions through SourceX's review and opens the rollout only when four feasibility gates pass
Pilot it in a handful of chapters or offices before the whole network hears about it. Draw six, give three the full enablement kit and three a link-only version, and after 120 days read two things: what members actually wrote to CEOs, coded against a fixed misstatement list, and whether their introductions were verified and are still open in SourceX's review. Roll out only if four feasibility gates pass. The pilot cannot forecast how many introductions the network will make, and no results exist yet.
Read → - The partner FAQ experiment that removes repeated questions: a before-and-after question count with a control code, an eight-answer FAQ that restates SourceX's published pages in your own words and adds no rule, and verified applications as the measure that decides
An FAQ can stop the same questions arriving from every CEO you contact, but only an honest count shows whether it did. Log and code every question for at least three equal periods, publish a dated FAQ whose answers restate SourceX's terms and fit page in your own words, then log three more periods with the same codes. Keep the FAQ if covered questions fall, a control code stays level and verified applications per contacted company do not drop. No partner or SourceX results for this study are published yet.
Read → - The partner marketing checklist before publishing a payout claim: a ten-line approval sheet that ties every number, verb and promise about money to a sentence SourceX published, says whose money it describes and keeps the conditions in view
A payout claim is any line in your marketing that tells a reader what money could follow an introduction: your reward, the company's payment, when either arrives or how likely it is. Before it goes out, tie it to a sentence SourceX has published, say whose money it means, keep the conditions in the same view and cut whatever SourceX has not published, such as typical amounts, company estimates or dates counted from the introduction. If the line names SourceX in public, get SourceX's written authorization first.
Read → - The partner onboarding experiment that measures understanding: nine scenario cards keyed to the payment trigger, credit and the CEO's decision, scored on day 0 and again on day 30
Measure whether new partners can apply the published terms to a situation, not whether they finished the course. Give every new partner nine short scenario cards, three each on the payment trigger, credit and who decides, scored against the exact wording of SourceX's terms. Compare two truthful onboarding formats by assigning cohorts before they start, and use the day-30 retest, on a parallel set of cards, as the primary measure. No pass rates exist yet, and this page invents none.
Read → - The partner opportunity beyond selling another software seat: an eight-box canvas that tests whether introducing customer CEOs helps the customer first
It is a motion that sells nothing your product does. A SaaS company can introduce a customer's CEO to SourceX, which licenses business records directly from that customer, while the vendor keeps its product relationship and never touches the records inside its platform. The idea only works for customers with 50+ full-time employees at peak, years of their own records and a CEO who wants the conversation. Test it on the customer-value canvas below, then run a small opt-in pilot.
Read → - The partner program retrospective after six months: a midpoint review of a one-year agreement that sets the record against month-zero expectations, rereads the terms you signed and picks a mode for the second half
Treat month six as the midpoint of a one-year Channel Partner Agreement that renews only with fresh signatures, not as a payday check. Set what you expected at month zero against the record: introductions by entry month, how many have passed their 90-day attribution window, hours spent, what clients said and any incidents. Reread the terms version now in force against the one you signed. Then pick a mode for months seven to twelve (active screening, on-request only or exit) and write down the evidence that would reverse it at renewal.
Read → - The partner quality scorecard that includes company feedback: score each introducer on fit, trust and follow-through, and let the CEO's account of the introduction act as a gate
A useful referral partner quality scorecard rates what each introducer controls, not how many logos they bring. Score three things from dated records: fit (did their unique companies pass the published fit facts at a metadata screen), trust (what the introduced CEOs say about how they were approached) and follow-through (did the partner log, record and close each introduction). Treat trust as a gate rather than an average: one credible report that a CEO never agreed to the introduction pauses the partner, however full their pipeline looks.
Read → - The partner social post that makes its incentive clear: material-connection disclosure examples for seven post formats, placed above the fold, carried by every image and repost, and cleared with SourceX before a post names it
A partner's social post makes its incentive clear when the line saying SourceX may pay the partner sits in the post's opening words, above any "see more" cut, in language a stranger understands, and travels with every image, video, reply and repost that carries the partner's link. Two things come first: SourceX's written authorization, since the disclosure itself names SourceX as the other side of your partner relationship and section 4.5 of the sample partner agreement restricts that without it, and a draft that names no company and implies no outcome.
Read → - The partner terms summary should point to the governing agreement: a summary-versus-contract mapping table that names, for each line of a program-at-a-glance box, the signed document that decides it for the person reading and the link label that takes them there
A program summary on a partner's website describes the arrangement; it never governs it. Each line should name the document that decides it for the person reading, and link there with a label saying what they will find. For a partner, the signed Channel Partner Agreement and its Schedule A prevail over the website terms on rate, base and cap. For a company, its own signed agreements with SourceX decide its payout and timing; the partner terms do not govern its payment. Conditions that change a money claim stay next to the claim, not behind the link.
Read → - The platform manager's handoff when taking leave: a continuity checklist that dates every 90-day window, limits what the cover may do and closes the asks that should not wait
An active referral keeps moving without its introducer: SourceX may deal with a referred company directly, and the partner need not join every call. What leave puts at risk is everything around it: companies asked but not yet introduced, 90-day windows that close while you are away, a notice inbox nobody reads and a colleague tempted to resubmit a company. Hand over one dated row per referral, name what the cover may and may not do, and stop starting new asks two weeks before you go.
Read → - The portfolio company CFO's first 90 days: cash, covenants, systems and records
A new portfolio company CFO should spend the first 90 days securing cash visibility and lender reporting, learning the sponsor's value creation plan, and building two documents that are easy to defer: a systems-and-archive map and a covenant check. Together they make later decisions on migrations, add-ons or a data license fast.
Read → - The portfolio referral screen before a founder email: an eight-line worksheet that names the operating entity, its controller and the CEO before anyone writes
Screen before you promote. A portfolio operations director should not email a founder about data licensing until an eight-line worksheet is filled: the legal entity that holds the records, who controls it today, whether it reached 50+ full-time employees at peak, how many years of its own records it keeps, where it operates and in what language, whether the records are its own, whether the CEO is the addressee, and whether SourceX already knows the company. Uncertain lines become questions, not promises.
Read → - The post-event research memo an investment team will read: an anonymous theme synthesis template that reports what founders asked, how widely and what is still open, and never which company might fit
A venture event insights memo tells the investment team what founders asked about licensing their companies' records for AI training, roughly how many raised each theme and which questions are still open. It carries no company's data, quotes nobody who has not approved the exact words in writing, leaves out anything said in confidence and never names a company that might fit. Build it from anonymous theme cards, print a theme only when three or more people raised it, and close with decisions the fund owns.
Read → - The post-transaction referral checklist for advisors: a six-sheet closing-to-introduction review pack that takes each closed deal in your book from the seat you held to ask now, relay, wait or drop
After closing, start from your own seat, not the company's records. Confirm the closing from public or company sources, note which side you advised and what your engagement letter still binds, and name today's decider: for an acquired company, the acquirer's CEO; for records a seller kept in an asset sale, the seller's CEO. Then check the company's new group, take fit facts only from that CEO and pick one route. Each closed deal ends as ask now, relay, wait or drop.
Read → - The private equity exit backlog in 2026: what operating partners can do while they wait
The private equity exit backlog in 2026 is roughly 32,000 unsold portfolio companies worth about $3.8 trillion, according to Bain, with buyout holds at exit near seven years. For operating partners, it shifts the value creation plan toward initiatives that pay during the hold, such as licensing a company's operating records, rather than only at exit.
Read → - The private equity fund lifecycle, stage by stage, and what operating teams do in each
A private equity fund lifecycle runs through five overlapping stages: fundraising, an investment period, the hold and value creation phase, a harvest period of exits and distributions, and extension or wind-down, commonly inside a ten-year term set by the LPA. Operating teams move from diligence and 100-day plans to cash, exit readiness and tail assets.
Read → - The private introduction after a public founder event: a consent transfer checklist that sorts what a founder said on stage from what was said around it, and shows which permissions carry into a private introduction and which the CEO must give again
A founder who talks publicly about the company's records, systems or AI plans has given the audience information, not permission. Remarks made on the record may be repeated with attribution, but they do not authorize an introduction, the reuse of a recording, or anything said in the hallway, at dinner or in a session held under the Chatham House Rule. Before a fund introduces the company to SourceX, its CEO has to agree to that specific introduction, and the company then tells SourceX about itself.
Read → - The professional referral audit trail without excess client detail: a nine-event decision register that records what was decided, when, by whom and with what outcome, points to where the proof already sits, and copies no client fact into the trail
Keep a record of decisions, not of the client. Each line in a professional referral audit trail should say what was decided, when, by whom and with what outcome, and then point to where the proof already lives inside the firm's own systems: a conflicts entry, counsel's memo, the CEO's email. It should never copy the client's facts, files or the reasons behind a confidential judgment. Log declines and decisions to stay out as carefully as introductions that go ahead.
Read → - The quarterly referral review that ends with a decision: find where companies are waiting, name who controls that stage, then continue, adjust or stop
A quarterly referral review is finished only when it records one decision: continue, adjust or stop. Write the decision rules before the quarter starts, count unique companies by stage with open cases kept in, then find the stage where most companies are waiting and ask who controls it. If your own team controls that stage, adjust. If SourceX or the Acquisition Partner does and fit-passing companies remain, continue. If no fit-passing companies are left to ask, or the hours exceed your ceiling, stop or pause.
Read → - The questions a founder should answer before being introduced: twelve, sorted by who can answer them, and the only two answers that stop an introduction
Give the founder twelve questions to answer privately before you introduce anyone, and say which answers matter. Only two answers stop an introduction: "I don't want this conversation" and "I'm not the CEO, and the CEO hasn't agreed." Everything else, from peak headcount to rights in customer contracts, can be answered roughly or with "I don't know", because SourceX's review settles those facts. Funding stage settles none of them. The founder then decides which answers travel with the introduction.
Read → - The questions business owners ask before joining the AI data economy: 20 questions sorted by the four decisions behind them, and who should answer each
Before an owner joins the AI data economy, their questions sort into four decisions: may we license these records, can we spare the effort and keep control, is it safe for our customers and staff, and is it worth it. A partner can answer the program questions from published pages. Questions about a specific dataset's rights, scope and value belong to SourceX's review, and contract questions belong to the owner's own counsel. Knowing which is which keeps the first conversation honest and short.
Read → - The questions fund counsel will ask about referral economics: a four-part briefing packet that answers ten money questions from SourceX's own documents, leaves the fund's papers to counsel and names the four points SourceX decides
Counsel will ask about the money before ruling on anything else: who would sign as partner, who pays the reward and from whose pocket, what it is calculated on, when it becomes payable, what can shrink it and who sits on the other side of the deal. SourceX's published terms and sample agreement answer most of that in their own words. What the fund's documents require, and four points the terms leave to SourceX, they cannot answer. The briefing packet below keeps those three sources apart.
Read →
Free resources
- Business succession planning assessment — Ten questions on successor, transition and documentation.
- NPV calculator — Net present value with a discounted cash flow table.
- Time value of money calculator — Future and present value with optional regular payments.
- All free tools · MCP resource center
Know a company with valuable proprietary data?
Become a referral partner, get your link and introduce the company’s CEO (in an owner-run company, usually the owner). Referral partners can join from most countries. We can't work with partners in countries or regions under U.S. sanctions or on U.S. restricted-party lists. The companies you introduce should be based in the U.S., Canada and the EU, or in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia with at least 70% of their data in English.
Refer a company →I own a business
Explore licensing your company's data directly to SourceX for AI training and evaluation. Start a short assessment; no uploads needed.
Start an assessment