This page is for partner growth managers deciding whether to add video to client outreach. Video feels personal and is easy to forward; text is quick to scan and quick to correct. Which one helps a CEO decide well is an empirical question, and the protocol below answers it for one firm's clients without inventing a winner in advance.
What a partner's video is for, beside SourceX's walkthrough
SourceX's how-it-works page already offers a walkthrough it describes as the whole process in under a minute, 57 seconds of real platform screens with labelled examples. That covers the company's side: applying, offers, agreements, delivery and payment. A partner has no reason to remake it.
What a partner can explain, and SourceX cannot, is the referral itself: who the partner is to this company, that SourceX may pay the partner, that the CEO alone decides, and what happens if the CEO says yes. That is the script under test. Both arms link SourceX's page, so its walkthrough is a constant, not a variable.
The format comparison protocol
Hypotheses, two-sided, written before recording. H0: leading with video makes no difference to understanding or to decided answers. H1: it changes understanding in either direction, and any gain in decisions counts only if understanding does not fall.
| Element | Video arm | Text arm | Parity rule |
|---|---|---|---|
| Script | Six fixed lines, spoken in about 60 to 75 seconds | The same six lines, word for word | One script file, frozen and dated |
| Disclosure | Shown on screen and spoken at the same moment | The same sentence in the body | Wording identical |
| Captions | Verbatim, with the speaker identified | Not applicable | Checked line by line before send |
| Visuals | The speaker on camera and plain title cards, no figures | None | No chart, logo, number or screenshot |
| Transcript | One click away, labeled read instead | The message is the transcript | Same text |
| Link to SourceX | How-it-works page | How-it-works page | Same link and position |
| The ask | One spoken line, repeated as text under the video | One line | Same words |
The six-line script.
- Who I am to you, and why I thought of your company.
- SourceX licenses a company's own operating records, with its agreement, for training and evaluating AI.
- You decide whether to talk to SourceX; nothing happens without your agreement.
- If you say yes, I introduce you and the company applies itself; SourceX decides fit in its own review.
- SourceX may pay me a reward separately; your company's payout is the same either way.
- Saying no is completely fine, and you can answer by replying to this message.
Line 5 rests on section 4 of the terms: SourceX pays the reward on its own account and never takes it out of what the company is due.
Two borrowed standards. W3C's guidance on captions for prerecorded video says captions carry the dialogue plus speaker identification and meaningful non-speech sounds; it exempts media that is a clearly labeled alternative to text (our paraphrase). Whether a partner's video would fall under that exemption does not matter here: the protocol captions it regardless, so the test compares formats rather than access. The FTC's Endorsement Guides describe a clear and conspicuous disclosure as difficult to miss and easily understood by ordinary consumers, ask for it in both the visual and audible portions when a message uses both, and want spoken disclosures at a volume, speed and cadence people can follow (our paraphrase). Neither text sets rules for referral videos; both set the bar this test holds itself to.
A recording freezes one terms version
Section 11 of SourceX's terms records changes taking effect in versions dated 2026-10-12, 2026-10-19, 2026-10-21, 2026-10-22 and 2026-10-22.2. A sentence in an email can be corrected in a minute after a change like that. A recording has to be shot again, and old copies keep circulating once forwarded.
The protocol handles this in three ways. The script holds no figures, rates, caps, deadlines or eligibility thresholds, only facts that have stayed stable. The file name carries the terms version checked on recording day. And on every new terms version, someone compares the six lines with the published text; if any line no longer holds, the video arm stops. A caption overlay that patches a spoken line does not count as a fix.
Assigning CEOs to a format, and what they consent to
Unit. One company, reached through one CEO the sender knows personally. Draw the list before recording, and split it at random within each sender's contacts, so every sender sends both formats.
Assignment is to the format the message leads with. A CEO in the video arm who reads the transcript instead stays in the video arm. Whether each CEO watched or read is recorded from the follow-up question and reported as a description only, because who chooses to watch is not random.
Consent. Nothing about the message announces a test; it is a normal ask from someone the CEO knows. Log each reply with its date. After the CEO decides, and only with their agreement, send three short statements to mark true, false or unsure, plus "Did you watch the video, read the text, or both?" Video hosts' viewing statistics are not used: they show plays, not understanding, and the test does not need them. Anyone who asks to be excluded is taken out, and the record keeps only that an exclusion happened.
What is scored: understanding first, then decisions
The three statements.
| Statement | Correct answer | What a wrong answer reveals |
|---|---|---|
| "The CEO decides whether the company talks to SourceX." | True | The format blurred who decides |
| "The partner may be paid by SourceX, and the company's payout is the same either way." | True | The disclosure did not land |
| "Receiving this message means the company is likely to be accepted." | False | The format implied fit |
| Measure | Definition | Denominator |
|---|---|---|
| Understanding (primary) | All three statements answered correctly | CEOs who answered |
| Decided answer (primary) | A written yes or no by day 21 | CEOs messaged in that arm |
| Agreed introduction | The CEO asks to be introduced, or says the company will apply | CEOs messaged |
| Access problem | A CEO reports being unable to play or follow the content | CEOs messaged |
| Objection | A complaint or a request to stop receiving messages | CEOs messaged |
| Partner hours | Time to script, record, caption, check and re-record | Per sender |
"Unsure" counts as not correct. Count wrong and unsure separately in the report; they mean different things.
When the protocol stops, and the rule that picks a format
Stop rules.
- A spoken line goes stale. A terms change, or a change to a page the script relies on, makes any line inaccurate: stop the video arm.
- Caption error. A caption found to differ from the voice track: withdraw the video until corrected and set aside its sends.
- Fit implied. Any CEO reads either format as saying the company qualifies or is endorsed: pause both and rewrite line 4.
- Access failure. One CEO reports being unable to reach the content: pause and fix the delivery before any further send.
- Objection. Pause and review.
- Planned end. Finish at whichever comes first: the CEO count or the date set before recording.
Choosing a format. Write the margins as numbers before launch. As an Illustrative form: adopt video only if, per 30 CEOs who answer the statements, it has no more wrong or unsure answers on statements 2 and 3 than text, at least as many CEOs scoring all three correct, and no more than two fewer decided answers per 30 CEOs messaged. Anything else is inconclusive, and text stays: it costs fewer partner hours and is corrected in minutes when the terms change.
Silence, declines and CEOs who never press play
Every CEO messaged stays in the denominator of the format they were assigned. A CEO who never opens the message counts as undecided. A decline is a decided answer. A video-arm CEO who only read the transcript is analyzed as video, because leading with video is the decision being tested.
Random assignment supports reading a clear difference as caused by the leading format, for this firm's clients and this script. The watched-or-read split is descriptive: CEOs who choose video may differ from those who skip it.
Reading an uncertain result and deciding what to change
With a few dozen CEOs, the likeliest result is that neither format is clearly better; the five-replies arithmetic shows why. An inconclusive result still answers the practical question: if video did not raise understanding, its extra hours and its upkeep after every terms change are not paid back. If one statement drew more wrong answers in one format, rewrite that line in both and retest.
Earlier understanding tests on this site, the founder comprehension pilot and the executive understanding interview, compared explanations rather than formats; their scoring keys pair well with the three statements. If the video is later cut from a longer recorded session, the webinar reuse checklist applies.
Report counts with denominators, every stop rule that fired, the terms version each recording was checked against, partner hours per format and what did not work, including lines CEOs misread.
Methods note. The protocol, script and measures were written on 2026-10-28 from SourceX's how-it-works page, program terms and screening page, with W3C's captions guidance and the FTC's Endorsement Guides definitions borrowed as standards. They describe a design. SourceX publishes no results from it.
What neither format can change
A format changes how a message lands, never who decides or whether a company fits. Whether to talk to SourceX is the CEO's call, and so is the signature. The published screen covers peak full-time headcount of 50 or more excluding contractors, records the company created over several years, licensing rights over them and fit with 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. Before submitting an introduction, the sender signs the SourceX Channel Partner Agreement.
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 recording anything, register, check line 4 of the script against who qualifies and line 5 against the current program terms, and note the terms version in the file name; SourceX's own process overview is the page both formats link.