Why a pile of questions is not yet a report
Most platform teams already hold the raw material: notes from calls, a shared inbox, perhaps a question backlog for founder questions that routes each one to the person who should answer it. A routing log answers who replies. A frequency report answers how often a kind of question comes up, and among whom, and a log built for routing gets that wrong in three predictable ways:
- One talkative founder dominates. Twelve questions from one CEO outweigh eleven founders who asked one each.
- Silent founders vanish. A founder who heard about SourceX and asked nothing leaves no row, so the share of founders with questions looks like everyone.
- Prompting is invisible. A question asked because the investor said "any questions?" is a different signal from one the founder raised unprompted.
The fix is a unit, a denominator and two flags: count founders, divide by every founder told, and mark each question for stage and prompting. What founders believe before anyone explains anything is a separate measurement, set out in what founders think data licensing means before the first call. This study starts after the investor has spoken.
The method in five stages
GAO's guide to content analysis organizes that kind of work in five steps: selecting the material, developing an analysis plan, coding, ensuring data reliability and analyzing the data (our summary of the guide's own description). This study follows the same order.
| Stage | What this study does | What it never does |
|---|---|---|
| Material | Uses capture cards written by participating investors | Uses forwarded emails, call recordings or company documents |
| Plan | Fixes the frame, window, families and measures before the first card | Adds a family after cards are read |
| Coding | Gives each question one primary code and up to two secondary codes | Codes the founder's motive or mood |
| Reliability | Has two coders code every card separately, then reports agreement by family | Lets the investor who wrote a card code it |
| Analysis | Reports founder-level shares with intervals, by stage and prompting | Ranks families by importance or links them to outcomes as causes |
Frame, consent and the capture card
Frame. Every founder or CEO a participating investor told about SourceX in one-to-one conversation or messages during two consecutive quarters. Newsletter mentions are out, because a question cannot be traced to a mention nobody remembers sending. Investors also keep a count of companies they considered and chose not to tell, since that choice shapes who enters the frame.
Consent. When a founder asks something, the investor asks whether a paraphrase with no company details may be logged for an aggregate study. A founder who says no stays in the frame as "told, not logged", so the denominator does not shrink to the agreeable.
What stays out. Anything the investor knows through a board seat or board materials. SourceX's own screening page notes that what a fund learns through board packs or information rights usually stays confidential; here it never reaches a card.
| Field | What to write | Never write |
|---|---|---|
| Date heard | The day of the conversation | A date reconstructed weeks later |
| Stage | First mention, after reading SourceX's pages, or after a conversation with SourceX | A guess at stage |
| Asker's role | Founder-CEO, other founder, CFO, other executive, board member | The person's name |
| Prompted | Yes if it answered "any questions?", no if raised unprompted | Nothing else |
| Question | A paraphrase of 25 words or fewer | Company name, figures, customers, systems |
| Answered how | Published SourceX page with the date read, routed to SourceX, routed to the company's advisers, or not answered | The investor's own view of fit, price or rights |
| Consent | Logged with agreement, or told but not logged | Anything for a founder who declined |
The question-frequency taxonomy
Eight families, each split into codes, plus two codes that keep the frame honest. The last column records whether SourceX's published pages answer the family today, which turns the report into a map of where founders ask questions the public pages leave open.
| Family | Codes | Example (Illustrative paraphrase) | Published answer today |
|---|---|---|---|
| F1 Fit | Size counting; years of records; location and language; whether the records are wanted | "Do contractors count toward the size line?" | Partly: the who-qualifies page sets 50+ full-time employees at peak, contractors excluded, and a geography sentence; relevance is decided in SourceX's review |
| F2 Ownership and clients | Customer-owned material; contracts limiting reuse; personal information | "Some files came from customers; are they ours to license?" | Partly: client-owned material and personal consumer data are usually not a fit; contracts go to the company's counsel and SourceX's review |
| F3 Exposure | Being identifiable; customers or rivals finding out; keeping the deal quiet | "Could anyone tell the records came from us?" | Not settled by the terms or the who-qualifies page; SourceX and the company's counsel answer it |
| F4 Effort | Who does the work; systems access; time | "How much of my team's time does this take?" | Asked of SourceX; investors do not answer it |
| F5 Money | Amount; timing; whether the investor's reward reduces the company's payout | "Does your reward come out of what we get?" | The third code only: section 4 of the terms has SourceX pay the reward itself, with nothing taken from the company's agreed compensation; amounts are not published |
| F6 Authority and exit | Who signs; the board's role; stopping | "Does the board have to approve?" | Partly: under the program the CEO opens a licensing discussion and signs; governance questions belong to the company |
| F7 Strategy | Helping AI that competes; exclusivity; reputation | "Are we training a competitor?" | Not on the program pages; asked of SourceX |
| F8 The investor's role | Whether the investor is paid; the fund's benefit; why this founder, now | "Why are you bringing this to me?" | The investor answers about themselves; the terms describe partner rewards |
| X Other | Anything that fits no family | Kept verbatim in paraphrase | Published as counterexamples |
| S Silent | Told, asked nothing | No card text | Counted in every denominator |
A question about SourceX as an investment is coded V and counted, then left out of the analysis: it belongs to a different decision, and the investor points the founder to the investor overview instead of answering.
Measures and the analysis plan
All shares divide by founders told in the window.
- Prevalence by family (primary). Founders who asked at least one question in the family, with a 95% interval.
- Silent share. Founders told who asked nothing.
- Stage and prompting. Prevalence at first mention against later stages, and unprompted against prompted.
- Coverage gap. The share of founders whose questions fall only in families with no published answer.
- Question counts. Published, but always labeled secondary and always beside the founder-level share.
- Outcome cross-tab. Ninety days after the window closes, families against each founder's last recorded stage (declined, no reply, agreed, applied), described and never read as cause.
No family figure is published from fewer than 30 founders told, and no cell in a cross-tab rests on fewer than 10 founders; smaller cells are folded into a broader row, never dropped. Which of those questions SourceX alone can settle is sorted in which eligibility questions belong with SourceX.
Founders who declined, went quiet or did not fit stay in
Every founder told keeps a row: those who declined, never replied, turned out not to fit, or refused logging. Their questions, or their silence, are part of the answer. Dropping them would describe only founders who stayed in the conversation, and that is the group least likely to have been put off by an unanswered question.
Three cards, coded (Illustrative)
| Card as first drafted | Card as kept | Codes |
|---|---|---|
| "Logistics CEO, 90 staff plus 40 contractors, asked if contractors count" | "Asks whether contractors count toward the size line" | F1 size counting; unprompted; first mention |
| "Asked whether our fund gets paid if they license" | "Asks whether the investor or the fund is paid" | F8 payment; secondary F5 |
| "Worried [customer name] would find out" | "Asks whether customers could learn of a license" | F3 customers finding out |
Each correction removes a fact the founder never agreed to share: a headcount, a contractor count, a customer's name. The question survives; the company does not.
Two working guesses, and the counts that would retire them
These are hypotheses to test, not findings.
- The investor's role comes up first. Founders ask about the investor's pay and motive (F8) at first mention more often than about fit (F1). Retired if F1 prevalence at first mention matches or exceeds F8's.
- Silence is not comfort. Founders who later declined were more often silent than founders who agreed. Retired if the silent share is no higher among decliners, which would suggest declines happen for reasons no question revealed.
The results table, empty until the window closes
| Family | Founders told who asked | At first mention | At later stages |
|---|---|---|---|
| F1 Fit | Not yet counted | Not yet counted | Not yet counted |
| F2 Ownership and clients | Not yet counted | Not yet counted | Not yet counted |
| F3 Exposure | Not yet counted | Not yet counted | Not yet counted |
| F4 Effort | Not yet counted | Not yet counted | Not yet counted |
| F5 Money | Not yet counted | Not yet counted | Not yet counted |
| F6 Authority and exit | Not yet counted | Not yet counted | Not yet counted |
| F7 Strategy | Not yet counted | Not yet counted | Not yet counted |
| F8 The investor's role | Not yet counted | Not yet counted | Not yet counted |
| S Silent | Not yet counted | Not yet counted | Not yet counted |
The release adds coder agreement by family, the number of investors and founders in the frame, the consent rate, the count of companies considered but not told, every X-coded question and the date each published answer was checked.
Where a question count misleads
- Founders choose what to ask an investor. Questions about money or control may go to a lawyer or a co-founder instead, so low prevalence can mean the question went elsewhere.
- Paraphrase drift. The card holds the investor's words, not the founder's. Writing it within one business day narrows the drift without removing it.
- Investors choose whom to tell. The frame reflects each investor's own screening, which is why the not-told count is published.
- Published answers move. The last column of the taxonomy is dated; when SourceX's pages change, the coverage gap changes with them.
- Frequency is not weight. A question asked by most founders may matter less to a decision than one asked by a few.
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 founder's questions end in a CEO who wants the conversation, register as a partner, answer only from who qualifies and the program terms, and let the company apply itself. Related study designs are collected in the guides library.
Counting founders' questions is research about your own network; whether SourceX is a company worth investing in is a separate question, and the SourceX investor overview takes requests for investor materials.
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.