# Venture network growth vs real data introductions | SourceX

Canonical: https://refer.sourcex.si/guides/where-a-growing-venture-network-creates-real-data-introductions
Last updated: 2026-11-05

A venture network creates data introductions only when its growth adds companies that are new to your register, pass the published fit facts and have a CEO who agrees to talk. Most added contacts do none of that: they point at companies already listed, companies below 50 full-time employees at peak, or companies nobody can ask. Count growth at the company level, follow each partner cohort through qualification and handoff, and test concentration before crediting the network. SourceX has not published cohort figures; this page publishes the method.

Funding status and network size never establish eligibility. A platform that doubles its founder community, LP roster and advisor bench has doubled the people it can reach. Whether it has added a single licensable company is a separate question, and the only one that matters for the referral program.

## Three statements on this page, and how sure each one is

| Statement | Kind of claim | How to check it |
|---|---|---|
| A company produces credit for one partner only, whoever introduced it validly first within the 90-day attribution window, and only after it submits a verified application | Program fact, terms Version 2026-10-22.2, since replaced by version 2026-11-04 | Section 2 of the [program terms](/terms) |
| If any earlier lead, account or email correspondence means SourceX already knew the company, it goes to duplicate review and no later partner receives credit | Program fact, same terms | Section 3 of the terms |
| Whether an introduced company qualifies, licenses records or produces any reward | Conditional | Decided only in SourceX's review and the company's signed agreements |
| As a network grows, authorized introductions grow far more slowly than contacts, because new contacts increasingly point at companies already listed | Our hypothesis | The cohort study below; no results exist yet |

## What does growing mean when you count companies instead of contacts?

Growth at the company-introduction level is the number of companies, new to your register, that reach an authorized introduction in a period. Everything else is reach.

| Unit | Definition used here | Counted when |
|---|---|---|
| Contact | A person someone in the network can reach | Added to any list |
| Introducer | A network member who can ask a CEO for a conversation: a founder, LP, co-investor, operator or advisor | Their first recorded ask |
| Company row | One operating business, recorded once, however many people know it | First appearance in the register |
| Authorized introduction | A company whose CEO knows about and wants the conversation, submitted by a partner who signed the SourceX Channel Partner Agreement | Submission |
| Qualified company | A company SourceX's review found eligible for a current program | Review outcome recorded |

**The four-bin ledger.** Every new contact lands in exactly one bin, and the bins must add up to the period's new contacts:

1. **Duplicate.** The contact works at, owns or advises a company already on your register.
2. **No fit.** A new company that misses any published fit fact: its full-time headcount never reached 50+ at peak (contractors do not count), its operating history is undocumented, it lacks rights to license, or it is based somewhere the program does not cover. 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.
3. **No path.** A new company that may fit, with nobody willing to ask its CEO.
4. **Authorized introduction.** The only bin that is growth for this purpose.

If the four bins do not sum to new contacts, the register is double counting somewhere, and no rate built on it can be trusted.

## Repeat introducers, new businesses and duplicated contacts send different signals

They look identical in a contact count and mean opposite things.

**Repeat introducers.** A portfolio founder who brings a third company is valuable, but that is depth in one relationship, not breadth in the network. Record an introducer code on every row so you can report what share of authorized introductions came from people who had introduced before.

**Genuinely new operating businesses.** A company is new when it is new to the register, not when a new person mentions it. The second contact at a listed company adds a relationship column to an existing row.

**Duplicated contacts.** Two people often know the same CEO: a portfolio founder and an LP may both know the same distributor. The program credits one partner per company and decides attribution from dated records, so a duplicate adds coordination work and the risk of two uncoordinated asks, never a second introduction. Agree inside the team who owns each relationship before anyone contacts the company; the comparison of [first valid referrer and last touch](/compare/referral-attribution-first-vs-last) explains why the order of asks matters.

The terms add a duplicate your register cannot see. SourceX checks its own records before inviting an introduced company, and any earlier lead, account or email correspondence counts as a relationship. Code those outcomes "known to SourceX" so a later analysis counts them neither as network failures nor as network successes.

## Track authorized introductions by partner cohort

Group introducers by the quarter of their first recorded ask, not by the date they joined a community or newsletter, since joining costs nothing. A cohort view shows whether newer introducers produce authorized introductions at the rate the first ones did, or whether recent growth is still carried by the founding group.

| Metric per cohort | Numerator | Denominator | What it tells you |
|---|---|---|---|
| Introductions per active introducer | Authorized introductions | Introducers with at least one ask | Productivity |
| New-company share | Asks that reached a company new to the register | All asks | Breadth |
| Duplicate share | Asks that hit a listed or SourceX-known company | All asks | Saturation |
| Repeat-introducer share | Authorized introductions from people with an earlier one | All authorized introductions | Depth |
| Concentration | Sum of each introducer's squared percentage share | Each introducer's share of the cohort's authorized introductions | Fragility |

The live guide to [tracking accepted introductions by network segment](/guides/how-to-track-accepted-introductions-by-professional-network-segment) covers the external log a partner keeps next to the dashboard; the cohort table adds the date of first ask so segments can be compared over time.

## Compare qualification and handoff completion as the network expands

After submission, a company passes stages SourceX controls: a verified application (the credit point), an eligible review outcome, a signed Confirmation under the company's licensing agreement, the Acquisition Partner's final written acceptance with completed handoff, and payment of the supplier. Report the share of each cohort reaching each stage within the same fixed window from its verified applications, for example 180 days. Without a fixed window the newest cohort always looks worst, because it has had the least time. Cases still open at the cutoff are reported as open, not as failures; the [conversion benchmark guide](/guides/data-referral-conversion-benchmark) explains that denominator and censoring logic for a single cohort, and this study repeats it across several.

## Three quarters of growth that added no introductions (Illustrative)

Illustrative. Invented counts for one fund's platform, built to show the ledger; not SourceX data, a benchmark or a forecast.

| Quarter | New contacts | Duplicates | No fit | No path | Authorized introductions |
|---|---|---|---|---|---|
| Q1 | 40 | 6 | 18 | 10 | 6 |
| Q2 | 70 | 19 | 30 | 14 | 7 |
| Q3 | 110 | 41 | 44 | 19 | 6 |

Reach almost tripled. Authorized introductions stayed between 6 and 7. The duplicate share rose from 15% to 37% of new contacts, the signature of a network that is growing around the same companies. Concentration moved too: Q1's six introductions came from six different introducers, an index of about 1,667 on the 10,000-point scale; Q3's six came from two people, four and two, an index of about 5,556. A platform reporting "network up 175%" would be describing the first column and hiding the last.

## Build the company-level network growth study

| Element | Rule |
|---|---|
| Sampling frame | Every company row on the register on the start date, plus every row added during the study; rows are closed with a reason, never deleted |
| Cohort dates | Introducer cohorts by calendar quarter of first ask; company cohorts by quarter of first appearance |
| Unit of analysis | The company; contacts and introducers are attributes of its row |
| Outcomes | Authorized introduction, verified application, eligible outcome, signed Confirmation, accepted handoff, supplier paid; each with a date and its evidence, such as the dashboard status or a written notice |
| Observation window | A fixed number of days from verified application, the same for every cohort; open cases reported separately |
| Anonymization review | Company codes, no names, no record contents; a second person checks that no row can be identified from industry, size and city before anything leaves the team |
| Reproducibility | Definitions, register snapshot dates and the analysis script filed with each report, so a rerun on the same snapshot gives the same numbers |
| Conclusion rule, stated in advance | "The network is creating more introductions" only if authorized introductions rise across three consecutive cohorts and the rise survives removing the single largest introducer |

**Method note.** The study measures counts, stages and concentration from the team's own register and from statuses SourceX reports to the partner, on dated snapshots. Demand for a company's records, how good those records are and any payout all stay outside the study, because only SourceX's review assesses them. Partners never request, receive or describe raw records; every field is metadata.

## Concentration and small samples: explanations that could reverse the conclusion

- **One introducer is the network.** The Justice Department's Antitrust Division describes the Herfindahl-Hirschman Index as the sum of each participant's squared share, near zero when many similar participants compete and 10,000 when one holds everything. Applied to introducers, a high index means the trend belongs to one or two relationships. Rerun every result without the top introducer.
- **A dozen cases cannot carry a rate.** Three qualified companies out of twelve applications is the greater of $50,000 plus 5% and 15%, but the exact 95% interval for that count runs from about 5.5% to 57.2%, using the binomial method the NIST handbook works through. Two cohorts at the greater of $50,000 plus 5% and 15% and 40% on samples that small are not shown to differ.
- **Depletion.** Early cohorts introduce the obvious companies first; later ones face a picked-over list. A falling rate can mean the network has outgrown the supply of fitting companies it can reach, not that new introducers are worse.
- **Program change.** Criteria belong to whichever program is running, and a company that misses one program can still qualify for the next. A cohort introduced during a narrow program will look weak for reasons outside the network.
- **Demand outside the network.** If synthetic or public data comes to cover more of what model developers need, demand for licensed business records could narrow, and every cohort would look weaker for a reason no network method controls.
- **Hidden duplicates.** Companies already known to SourceX are not credited. If one cohort happens to hit many of them, it will look weaker at finding companies than it was.
- **Borrowed attribution.** Partner programs often measure loosely: when PartnerStack and Wynter surveyed 100 senior leaders at B2B SaaS companies with more than $50M in revenue, 42% said they relied on multi-touch attribution, 31% on first-touch and 19% on last-touch, while 8% used another method or tracked nothing. A study that inherits someone else's attribution inherits its gaps; keep your own dated register.

**A measurable next step.** Rebuild last quarter's growth as a four-bin ledger from the register you already have. If authorized introductions are under a tenth of new contacts, spend the next quarter on paths to CEOs at listed companies rather than on adding names. To test whether a narrower list does better, the [ten-introduction experiment](/guides/the-ten-introduction-experiment-for-an-ai-data-network-thesis) sets decision rules before the first ask, and the [five-ring map](/guides/your-next-ai-data-introduction-may-sit-outside-the-portfolio) shows where in the network the next rows usually come from. Everything else about the program is in the [overview for venture platform teams](/for/venture-platform-teams) and the [guides hub](/guides).

## 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 a row in the fourth bin has a CEO who wants the conversation, [register for a partner link](/register), hold the company up to [who qualifies](/who-qualifies) and study the [program terms](/terms), above all the duplicate rule that settles whether a new company really is new.

Readers weighing SourceX itself as an investment face a separate decision from any referral, and can request investor materials through 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://www.justice.gov/atr/herfindahl-hirschman-index
- https://www.itl.nist.gov/div898/handbook/prc/section2/prc241.htm
- https://partnerstack.com/resources/research-lab/the-state-of-partnerships-in-gtm-2026
- https://sourcex.si/investors