# Sector brief vs general explanation: a test | SourceX

Canonical: https://refer.sourcex.si/guides/test-a-sector-specific-brief-against-a-general-explanation
Last updated: 2026-11-04

Test the specificity of one paragraph, not a new brief. Write a single introduction in two versions that are identical except for the paragraph naming the kinds of records a company might hold: the general version uses SourceX's screening examples, the sector version uses the examples SourceX publishes for that company's industry. Split CEOs at random within each sector, read agreed introductions beside corrections and misreadings, and stop if any sector line cannot be traced to SourceX's own page. This page is the design; no results are published here.

It is written for growth leads on venture platform teams who suspect a founder answers faster when the records named sound like their own business. The suspicion is reasonable and untested. A sector paragraph can also misfire: it names records a company does not keep, or reads as a hint that SourceX is looking for that industry. The test measures both.

## The one paragraph a sector brief is allowed to change

**Question.** When everything else in an introduction stays fixed, does naming record examples from the company's own sector change how many CEOs agree to an introduction, and how many reply that the examples do not describe them?

**Hypotheses, two-sided, written before launch.** H0: the sector paragraph makes no difference to agreed introductions, corrections or misreadings. H1: it changes agreed introductions in either direction, and any gain counts only if corrections and misreadings do not rise past a margin set in advance.

**What stays frozen.** Sender, subject line, greeting, the description of SourceX, the disclosure, the fit sentence, links, the ask and the send window. The two paragraphs differ in length by no more than a tenth, so the comparison is about specificity rather than length.

Earlier tests on this site change other things. [The two-angle worksheet](/guides/how-to-compare-two-outreach-angles-without-changing-everything) swaps the opening reason for writing; [the matched network study](/guides/do-specialized-networks-produce-better-qualified-referrals) compares specialist networks with general ones. Here the network, the sender and the reason stay the same, and only the record examples move.

## The sector relevance message test

Each row is one part of the message. Only the first row differs between versions.

| Part of the message | General version | Sector version (Illustrative: logistics) | Rule both obey |
|---|---|---|---|
| Record examples | Operating records a company created itself, such as SOPs, internal knowledge bases, project histories, QA records and support workflows | Operating records such as shipment records, dispatch communications, exception notes, freight documents and rate quotes | Every noun appears on the named SourceX page on the audit date |
| Where the examples come from | "These are examples SourceX gives on its screening page" | "These are examples SourceX lists on its logistics page" | One attribution line, same length |
| Rights | Records the company can license without breaching client, employee or third-party obligations | The same sentence | Sector considerations go in the follow-up, never the first message |
| Fit | SourceX checks in its own review whether the records match a current program | The same sentence | No sentence about demand for any sector |
| Disclosure | If your company goes ahead and SourceX pays it, SourceX may pay me a reward separately; your payout would not change | The same sentence | Rests on section 4 of the terms |
| The ask | Would you like me to introduce you to SourceX? | The same sentence | One ask, no deadline |

**The traceability rule.** A sector sentence may contain only record types and considerations that SourceX's industry page for that sector states on the day of the audit. SourceX publishes one such page per industry, from logistics and construction to insurance, accounting and claims administration. The logistics page, for example, also names shipper and consignee confidentiality, driver personal and location data, and carrier contracts as things to check; those stay out of the first message in both versions and go into any follow-up the CEO asks for.

**Why demand sentences are excluded.** SourceX's screening page says its current data needs differ by data category and that a company missing one program may fit another later, and its how-it-works page labels its fit illustration "Not a valuation", tying final value to actual Acquisition Partner demand and due diligence. A line such as "your sector is in demand" would claim something about this company that nobody can know before review, and would add a second difference between the versions.

## Drawing the sample and assigning each company a sector

**Unit.** One company, reached through one CEO whom the sender knows personally. Draw the full list before writing either version.

**Sector assignment.** A colleague who sends nothing assigns each company one sector before the draw, from its primary activity, and notes its NAICS sector code as a cross-check. NAICS, the classification federal statistical agencies use for business establishments, groups them by production process, and the first two digits of a code designate the sector (our paraphrase of the Census Bureau's description). A company whose sector has no SourceX industry page receives the general version and is reported apart, outside the comparison.

**Expect mismatches.** Industry codes describe what a company sells, not every workflow it runs. [The sector specialist guide](/guides/how-a-sector-specialist-finds-unusual-referral-candidates) shows how the strongest records can sit in a workflow filed under another industry. In this test that is a measured outcome: a CEO who replies "that is not what we keep" is logged as a correction, and nobody switches the company to a better-fitting version afterward.

**Randomization.** Split at random within each sector and within each sender, so every sender sends both versions in every sector they cover. A sector with fewer than four companies is described, not compared.

**Consent and what is recorded.** The message is an ordinary personal ask. The test records the reply, its date and two follow-up answers from CEOs who agree to give them. Nobody asks a CEO for headcount, systems or files to score the test, nothing beyond the introduction itself reaches SourceX, and a CEO who asks to be left out is removed, with only the removal counted.

## What is counted in each sector

| Measure | Definition | Denominator |
|---|---|---|
| Agreed introduction (primary) | The CEO asks in writing to be introduced, or says the company will apply | CEOs messaged in that version and sector |
| Correction | The CEO says the named examples do not describe the company's records | CEOs messaged |
| Sector read as approval | A reply reading the brief as saying the company qualifies, that SourceX wants its sector, or that the fund has arranged access | CEOs messaged |
| Wrong-fit request | A CEO asks to proceed while stating a fact the fit sentence rules out | CEOs messaged |
| Objection | A complaint, or a request to stop receiving messages | CEOs messaged |

After the CEO's decision, and only with their agreement, ask two things: "Did the examples match the records your company keeps?" and "Did anything suggest your company had already been chosen?" Opens and clicks are not measured. The terms do not let a link click alone decide credit, and a click says nothing about understanding.

## Stop rules and margins for a sector brief

1. **Untraceable line.** A sent sector paragraph names a record type the SourceX page does not list: halt that sector, set its sends aside and restart only after a fresh audit.
2. **The industry page changes.** If SourceX edits the page mid-test, the sector version has changed too: halt that sector.
3. **Approval read.** One reply reading either version as acceptance, endorsement or special access: pause both versions until someone outside the test rereads them.
4. **Objection.** Pause and review before any further send.
5. **Limit.** Close the sector at the CEO count and end date fixed before launch.

**Decision rule, per sector.** As an Illustrative form: adopt the sector version in that sector only if it shows at least three more agreed introductions per 30 CEOs, no more than one extra correction per 30, and no approval reads at all. Adopt the general version on the mirror result. Anything else is inconclusive: keep the general version, which needs no per-sector upkeep. A verdict in logistics says nothing about accounting.

## Missing, declined and unsuccessful companies stay in the count

Every company drawn stays in the version it was assigned. A CEO who never replies is in the denominator. A CEO who declines is a decided answer, not a failure of the message. A company that applies and is then declined in SourceX's review still counts as an agreed introduction, because the message did its job and fit is decided elsewhere.

Within one sector, a randomized comparison can support a causal reading for that network, that period and that wording. Differences between sectors are descriptive only: sectors differ in company age, size and relationship mix, and none of that was randomized.

## Reading a small result without overclaiming

Most platform teams can reach a dozen CEOs per sector, not hundreds. At that size a gap of two or three introductions is well inside chance, as [the five-replies arithmetic](/guides/how-to-avoid-declaring-a-winner-from-five-replies) shows, and [the ten-introduction experiment](/guides/the-ten-introduction-experiment-for-an-ai-data-network-thesis) explains what a tiny cohort can still reject. Publish per sector, with denominators:

| Line | What to report |
|---|---|
| 1 | Companies drawn, messaged and excluded per sector and version, with reasons |
| 2 | Agreed introductions, corrections and approval reads per version |
| 3 | Wrong-fit requests and objections |
| 4 | Audit dates, and any industry page change during the test |
| 5 | Stop rules that fired, and what was changed |
| 6 | The verdict per sector against the margins set in advance |
| 7 | What did not work: examples that confused, sectors where corrections ran high |

## What can make this design mislead

- **Programs move.** Data needs differ by category and change over time, so a sector paragraph that tested well can become misleading without a word changing.
- **Codes miss workflows.** Corrections may reflect sector assignment rather than the paragraph. Report them both ways.
- **Sender warmth.** Senders who know a sector well may add warmth in replies. Fix the text of the first message and log every follow-up.
- **Contamination.** Founders in one sector talk to each other; note any CEO who saw both versions and keep them in their assigned arm.
- **Self-report.** The follow-up answers come from CEOs willing to give them.
- **No benchmark.** Nobody has published sector brief results for data licensing introductions.

**Methods note.** The paragraph pair, frame, measures and rules were built on 2026-10-28 from SourceX's screening page, its industries index and logistics page, its how-it-works page and the program terms, with the Census Bureau's NAICS description used for sector assignment. They describe a design, not measured results.

## Who decides, and what the brief never collects

Whichever paragraph a CEO reads, the decision to talk to SourceX, and the signature after it, belong to that CEO. Fit rests on the published profile: 50 or more full-time employees at peak with contractors excluded, several years of the company's own records, rights to license them, data never licensed for AI training before, and relevance to a current program. 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. Senders sign the SourceX Channel Partner Agreement before their first introduction goes in, and every introduction names one company whose CEO said yes to it.

## 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.

Before drafting either paragraph, [register](/register), copy the general examples from [who qualifies](/who-qualifies) and the sector examples from SourceX's industry page on the same audit date, and read the [program terms](/terms) the disclosure line rests on.

Running a message test inside a network is referral work. Weighing SourceX as a company to back is a separate exercise, and readers doing that can ask for 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/who-qualifies
- https://sourcex.si/industries
- https://sourcex.si/industries/logistics
- https://sourcex.si/how-it-works
- https://www.census.gov/naics/
- https://refer.sourcex.si/terms
- https://sourcex.si/investors
