# Specialist network referral quality: a 90-day test | SourceX

Canonical: https://refer.sourcex.si/guides/the-small-network-advantage-in-specialized-data-referrals
Last updated: 2026-11-04

Referral quality in a specialist network comes from members who know which companies in their sector keep long operating records and who can ask those companies' CEOs directly, not from the network's size. That advantage is plausible but unproven: nobody has published comparison data for company data introductions. Before you reorganize a platform around it, brief a small specialist group and a broad community with the same page, count both from the same denominator, and judge them by authorized introductions rather than replies.

The asset is never the membership list. It is an authorized relationship with a company's CEO, held by a named person, at a company that passes the published fit facts. Portfolio membership, community membership and sector expertise are only ways of finding those relationships.

## The program facts, and the claim this page tests

Some things are fixed by SourceX's program terms and who-qualifies page, checked 2026-10-26. An introduced company needs 50 or more full-time employees at peak, documented operating history, rights to license and an authorized contact; under the program the CEO authorizes. The first valid introduction within the 90-day attribution window is credited once the company submits a verified application, one partner per company, and companies SourceX already knows are reviewed as duplicates. Whether a company's records match SourceX's current data needs is decided only in review.

The claim being tested is ours, not SourceX's: that a small specialist group produces more authorized introductions per member briefed than a broad community does. It is a working hypothesis with no published data behind it. The rest of this page is a way to find out whether it holds for your network.

## Why depth could beat reach, and why it might not

Four mechanisms favor the specialist group.

1. **Record literacy.** A former operator in specialty distribution knows which systems the sector runs, how far back order and exception histories usually go and which firms still keep them.
2. **Relationship depth.** Specialists often worked for, supplied or sat on boards of the companies they name, so the relationship owner and the person speaking are the same.
3. **Rights instinct.** People inside a sector know where records belong to clients, as in many agency and outsourcing models, and drop those companies before anyone asks a CEO.
4. **Less noise for the CEO.** A request from a respected peer about the company's own records reads differently from a forwarded program link.

Reach has its own case. Large platforms are built for breadth: Andreessen Horowitz, for example, describes its operating teams, spanning areas from marketing and talent to legal and policy, as the largest team of operators in venture. A network like that touches sectors no specialist group sees. Sociology adds a sharper objection. In "The Strength of Weak Ties" (1973), Mark Granovetter argued that the overlap between two people's networks rises with the strength of their tie, and that weak ties matter for the spread of information and for opportunity; his evidence came from how job seekers found work. Applied here, a tight specialist circle may keep rediscovering the same well-known firms, while loose ties in a broad community surface companies nobody in the circle knows.

Both stories are reasonable. Only a fair test separates them.

## Start from company and owner pairs, not members

The unit you count is a pair: one company and the one person who owns the relationship with its CEO. Members are only where pairs come from. For each pair, record the company, its CEO, the relationship owner, how the owner knows the CEO (worked there, board seat, supplier, customer, peer), which arm surfaced it, which fit facts are known and whether the CEO knows about the conversation.

Two rules keep the count honest. A pair without a named relationship owner is not reachable, however famous the company. A company named in both arms counts once, credited to whichever arm first produced an authorized path, the same logic as the program's first-valid-introduction rule. The pack guide on [counting one company across three networks](/guides/how-to-avoid-counting-the-same-company-across-three-networks) shows how to merge those rows.

## The specialization-versus-scale test design

| Element | Specialist arm | Broad arm | Why it is set this way |
|---|---|---|---|
| Who is briefed | 30 members of one sector group, such as former operators in specialty distribution | 30 members drawn at random from your wider community | Equal numbers keep effort comparable; a random draw stops you picking your keenest generalists |
| What they receive | One-page brief: fit facts, what partners never handle, ask the relationship owner first | The identical brief | A better brief in one arm would decide the result |
| Window | 90 days from the brief | 90 days from the brief | Matches the program's 90-day attribution window |
| Primary measure | Authorized introductions per 100 members briefed | Authorized introductions per 100 members briefed | Counts what the program can credit, not activity |
| Secondary measures | Pairs named, owners confirmed, fit facts complete, duplicates, no-fit reasons, member minutes | The same set | Shows where each arm loses rows |
| Not measured | Licensing outcomes and payouts | Licensing outcomes and payouts | Too slow, too few and not published; they would reward luck |
| Decision rule | Wins only with at least twice the broad arm's rate and at least five authorized introductions | Wins or ties otherwise | Fixed before the brief goes out, so a small lucky count cannot decide |

**What counts as an authorized introduction.** All four conditions hold: a named relationship owner has agreed to ask; the CEO knows about the conversation and wants it; all four fit facts are known and pass on the owner's knowledge; and nobody knows of an existing application or relationship between the company and SourceX. A verified application by the company is the lagging check, since that is when the program credits an introduction.

**Methods note.** The design is built from the program's published fit facts, attribution window and duplicate rule (SourceX terms and who-qualifies page, checked 2026-10-26). It measures introductions, not demand for any company's records, record quality or payout, which only SourceX's review and a signed agreement establish. Thirty members per arm can detect only large differences, which is why the rule demands at least twice the rate and a minimum of five. If neither arm reaches five, the test is inconclusive, and that is itself a finding. No results from this design have been published.

## Three introductions on the ledger (Illustrative)

Illustrative. Generic companies, invented facts, no outcomes implied.

| Row | Arm | Company | Relationship owner | Fit facts known | CEO wants the call | Counts |
|---|---|---|---|---|---|---|
| 1 | Specialist | 210-person specialty chemicals distributor, founder-led | Group member who ran its operations for six years | 4 of 4 | Yes | Yes |
| 2 | Broad | 450-person freight brokerage | Community member who is CFO of its largest carrier | 3 of 4 | Pending | Not yet |
| 3 | Specialist | 38-person testing laboratory, well known in the sector | Group member on its advisory board | Fails size | Yes | No, logged as no-fit |

**Row 1, the owner-led company.** The founder still runs the distributor as CEO. The group member who ran its operations knows it keeps fifteen years of order exceptions and supplier claims in one system, asks the founder directly, and the founder wants the conversation. Every condition holds on the first call.

**Row 2, the network intermediary.** The broad-arm member does not know the brokerage's CEO. They know its COO through the carrier relationship, and the COO agrees to raise it with the CEO. Until the CEO says yes, the row is reachable but not authorized, and the arm gets no credit. This is how many broad-network rows look: a real path, one link longer.

**Row 3, the no-fit.** The laboratory is the most respected name in the sector and the relationship is warm, but it has 38 full-time employees and much of its work product belongs to the clients who commissioned the tests. A specialist arm that counted fame would score this row; the ledger logs it with its reason and moves on.

## A reachable name is not a mandate-fit company

Specialist groups inflate reach as much as broad ones do: everyone in a sector feels they know its famous firms. The ledger keeps two columns apart. Reachable means a named relationship owner exists. Mandate-fit means the company passes the fit facts and has rights to license its own records. Location is one of those facts: 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. If the specialist arm works in staffing, only a firm's own full-time employees count toward the 50; workers placed at clients do not, and a firm with strong records is worth introducing anyway with both numbers stated.

Every question in either arm stays at metadata level: what records exist, roughly how many, over how many years, in which system and who owns them. Nobody in either arm asks for files, samples or logins. For a sector example of the same discipline, see how [accounting firm niche practices](/guides/accounting-firm-niche-industry-practice) spot deep records.

## Ordering the next conversations after the test

Rank what the test surfaced by evidence you can observe, never by projected proceeds; SourceX has not published conversion rates or typical payouts. Use these keys in order: an authorized path, all fit facts known, a dated event such as a documented system migration or archive deletion, a completed duplicate check, and the effort the relationship owner must spend. A no-fit row sits below every other row, however warm.

Then act on the result. If the specialist arm wins under the rule, recruit a second specialist group in an adjacent sector and run the test again before scaling. If the broad arm wins or ties, keep briefing broadly and use specialists as reviewers of rows rather than sources of them. The live guide to [measuring referral quality instead of volume](/guides/how-to-measure-referral-quality-instead-of-referral-volume) and the one on [prioritizing a network of owner relationships](/guides/how-to-prioritize-a-network-of-business-owner-relationships) cover the steady-state routine.

## Where the small-network thesis could be wrong

- **Weak ties.** Granovetter's argument predicts the opposite of this page's hypothesis for novel opportunities. If the broad arm names more companies nobody else knew, the weak-tie effect is winning.
- **Saturation.** A specialist circle may name the same thirty companies, and only one partner is credited per company. Duplicates that pile up in one arm are a cost of depth.
- **Demand.** SourceX's programs differ by company size, operating history, geography, language and data category, and relevance is checked in review. A sector the specialists know well may not match current needs at all.
- **Small numbers.** Five authorized introductions is a floor, not proof. Repeat the test before restructuring a platform.
- **Enthusiasm bias.** Members chosen because they like the platform lead may report "the CEO wants the call" too early. Confirm with the relationship owner in writing.
- **Relationship cost.** Every ask spends goodwill, and no data on how sector peers react has been published. Record declines with reasons.

## 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.

When either arm produces a row with a named relationship owner, a CEO who wants the call and four known fit facts, [register for a partner link](/register), check the row against [who qualifies](/who-qualifies) and read the [program terms](/terms); whoever submits signs the SourceX Channel Partner Agreement first. For the platform side of the program, start with the [overview for venture platform teams](/for/venture-platform-teams).

Evaluating SourceX as a company is a separate decision from running this test or referring anyone; for that question, ask for investor materials via the [SourceX investor overview](https://sourcex.si/investors).

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.

## Sources

- https://refer.sourcex.si/terms
- https://refer.sourcex.si/who-qualifies
- https://a16z.com/about/
- https://www.journals.uchicago.edu/doi/10.1086/225469
- https://sourcex.si/investors
