An adjacency is a company one step away from a portfolio company: a supplier it buys from, a customer it sells to, or another client of an advisor it uses. None of these relationships belong to the fund, and that is what the matrix makes visible.
Who owns each adjacency, and what does asking cost them?
The founder owns supplier and customer relationships; the advisor owns its client relationships; the platform team owns none of them. Its job is to explain the fit tests to the person who does, then let that person decide.
Suppliers. When a portfolio founder mentions data licensing to a supplier's CEO, a customer is raising something with a vendor. Vendors generally welcome a customer's attention, so the founder spends little. This is our working thesis about relationship cost, not a measured result.
Customers. The direction reverses. A founder who raises an unrelated topic with a customer's CEO spends goodwill with the account that pays them. Some founders will gladly do it; many will not, and the matrix should make declining easy.
Advisors' clients. The accountants, law firms, consultants and managed service providers that serve a portfolio company serve many other mid-sized companies too. Those clients are the advisor's relationships, held under client confidentiality and, for some professions, explicit disclosure rules. Platform programs already maintain such networks: a16z's About page, for example, describes a network of experts and an operating team spanning marketing, talent, legal and policy. The right first conversation is with the advisor about the advisor's own participation, never a request for names.
There is also a line on what the fund may know. Investor agreements commonly restrict how investors use a company's confidential information. One filed with Backblaze's 2021 registration statement, for example, has each investor agree to keep confidential, and not disclose, divulge or use for any purpose other than monitoring its investment, the company's confidential information it obtains. Supplier contracts and customer lists seen in board decks fall squarely inside that kind of clause. So the matrix starts from the adjacencies a founder chooses to name, plus public facts, and never from the data room.
The supplier-customer-advisor adjacency matrix
Each adjacency passes one gate, then takes three scores from 0 to 2.
The boundary gate. If the only way you know the company or its fit is through board or investor materials, stop. The row does not go on the matrix until the founder names it.
| Test | 0 | 1 | 2 |
|---|---|---|---|
| Owner willing | No one owns the CEO relationship | An owner is named but has not offered | The owner offered to ask |
| Cost of asking | Owner depends on the company (a key customer) | Neutral, or a professional duty applies | Owner is the company's customer, or the company is the owner's own firm |
| Fit visible | No fit facts known | One or two of the four known | Three or four known from public or volunteered facts |
Six points means a first conversation with the owner this month. Three to five means one more question to the owner. Zero to two means park the row.
How the three adjacency types usually start:
| Adjacency | Who owns the CEO relationship | Usual cost of asking | What you may know without board material | First conversation |
|---|---|---|---|---|
| Supplier to a portfolio company | The founder, as its customer | Low | Public headcount band, founding year, what it makes | The founder: would you mention it to them? |
| Customer of a portfolio company | The founder or the account lead | High | Public facts, plus anything the founder volunteers | The founder: would raising this help or hurt the account? |
| Another client of a portfolio company's advisor | The advisor | Professional duties apply | The advisor's own firm, never its client list | The advisor: does your firm want to refer its own clients? |
Method note. The tests come from three things this page can cite: who signs (each company signs its own agreement, per SourceX's portfolio guidance), what an investor may use (the confidentiality clause above) and the published fit criteria (SourceX terms and who-qualifies page, checked 2026-10-26). The weights are equal because no outcome data exists to justify others. Once you have recorded a few dozen first conversations, reweight from your own results. The matrix does not measure record quality, demand or payout.
Three adjacencies scored (Illustrative)
Illustrative. The three adjacencies are generic with invented facts, and their scores imply no outcome.
An owner-led supplier. A 230-person contract packaging company, still owned and run by its founder as CEO, packs products for a portfolio consumer brand. The portfolio founder has met that CEO every quarter for five years and offers to mention the program (owner 2). A customer asking a vendor (cost 2). Its website gives the headcount band and a 2004 founding date, and the portfolio founder knows it keeps years of batch and quality records in one system; rights are unknown (fit 1). Score 5: one more question, which the founder can ask in the same call.
A network intermediary. A 140-person regional accounting firm closes the books for three portfolio companies. The platform lead's instinct is to ask the firm which other clients might fit. That is the wrong conversation: those clients are the firm's relationships, and accountants may face disclosure rules on referral compensation, as the guide on disclosing a referral fee explains. The right conversation is with the managing partner, about whether the firm wants to register and introduce its own clients, and about whether the firm itself, with its own long operating history, is a candidate. As a platform-team row it scores 2 and is parked; as the firm's own pipeline it is not the fund's to score.
A no-fit counterexample. A portfolio software company's largest customer is a 2,000-person hospital network: large, decades old and well documented. It looks like the best row on the sheet. But the founder depends on the account (cost 0), has not offered (owner 1), and the most visible records are patient records, a pattern SourceX's poor-fit guidance flags. Size passes; rights do not look promising (fit 1). Score 2: park it, and do not raise it with the founder again unless they bring it up.
| Adjacency (Illustrative) | Owner willing | Cost of asking | Fit visible | Score | Verdict |
|---|---|---|---|---|---|
| Contract packager, supplier | 2 | 2 | 1 | 5 | One more question to the founder |
| Accounting firm's clients | 1 | 1 | 0 | 2 | Route to the advisor |
| Hospital network, customer | 1 | 0 | 1 | 2 | Park |
A cheap conversation still has to pass the fit facts
A high score says a conversation is cheap to start. Fit decides whether it is worth having. Four published facts set the bar: a peak of at least 50 full-time employees, with contractors left out; an operating history the company has documented; the right to license the records; and a base in an eligible jurisdiction. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 70% of the data is in English.
Because the adjacency's CEO is the person who approves a licensing discussion and signs for the company, every row needs a path that ends with that CEO, not merely a friendly executive. Questions stay at metadata level: record types, rough volume, years covered, the system and who administers it. Nobody asks for files, samples or logins. SourceX's review, not the matrix, settles whether the records match current demand, whether the company's own customer contracts limit reuse and what any license would cover. Two suppliers inside one corporate group make a single row, since the once-per-company floor treats a parent and its subsidiaries as one company; the guide on counting a company once across networks shows how to catch them.
Rank next conversations without projecting licensing proceeds
SourceX has not published conversion rates or typical payouts, so do not rank by guessed deal value. Rank by:
- Matrix score, highest first.
- A dated event at the adjacent company, such as a documented system migration or archive deletion.
- Founder load. Group every adjacency one founder owns into a single conversation, and set a limit on how often you ask any founder. A founder asked about six suppliers in six separate messages will stop answering the seventh.
Before any introduction is submitted, agree in writing who the partner is: the founder, the advisor or the platform team. The partner who submits must have signed the SourceX Channel Partner Agreement beforehand (copies signed back when its title was Channel Partner Non-Circumvention and Non-Disclosure Agreement keep their force as signed), and only one partner is credited for each company, the author of its earliest valid introduction, provided the company files a verified application while the 90-day attribution window is still open. For the mechanics of the founder-to-CEO step, see introducing a portfolio company CEO to a vendor.
Where this matrix is wrong
- Supplier power. A portfolio company that depends on a sole-source supplier has less leverage than the matrix assumes; score that supplier like a customer.
- The portfolio company inside the supplier's records. A supplier's order and quality history includes records about serving the portfolio company. A founder may reasonably ask how those would be treated. That is a question for the supplier's contracts and SourceX's rights review, and the founder should hear it before agreeing to ask.
- Confidentiality. Clauses like the one above vary by agreement. If in doubt about whether you learned something in confidence, treat it as confidential.
- Demand. SourceX checks relevance to its current data needs in review, so a six-point row can still be declined.
- Equal weights. Untested. If your recorded results show that "fit visible" predicts outcomes and "cost of asking" does not, change the weights.
A measurable next step. Ask three portfolio founders to name up to ten suppliers each, score them, and record for each row whether the founder asked, whether the CEO agreed to listen and how many fit facts passed. Compare the supplier rows with any customer rows a founder volunteered. If customers reach a CEO yes as often as suppliers, the cost-of-asking thesis is wrong for your portfolio.
Next step
Refer a company. No cap on rewards. Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.
Once an adjacency scores six points and its CEO has asked for the conversation, register for a partner link, test the company on who qualifies and review the program terms before anyone submits. For portfolio companies themselves, SourceX explains how a fund can coordinate reviews across companies in licensing data across a portfolio of companies, and the overview for venture platform teams covers the referral side. The guides hub lists related methods, including the consent audit for reachable decision-makers.
Weighing SourceX as somewhere to invest has nothing to do with scoring adjacencies; for that decision, requests for investor materials go through the SourceX investor overview.
This article is for information only and is not an offer to sell or a solicitation of an offer to buy securities. Any offering by SourceX is made only to verified accredited investors through its offering documents.