# Company referral stall reasons: a study design | SourceX

Canonical: https://refer.sourcex.si/guides/the-most-common-reasons-company-introductions-stall
Last updated: 2026-11-04

No published list of stall reasons exists for SourceX introductions, and this page does not offer one. It specifies an observational study that would produce the evidence: a cohort of introductions followed from the CEO's first yes, a clock at each stage that marks a stall only after a fixed number of days without a dated event, a reason codebook applied by two coders who cannot see outcomes, and censoring for introductions still open on the read date. Until the cohort closes and is analysed, there are no results to report.

The study would need data held by SourceX and by partners who consent, linked by introduction. Partners never see company records at any point, and the study adds no follow-up to any company.

## A stall is a missing event, not a decision

Introductions end in a handful of ways that someone decides: the CEO says no, SourceX reaches an outcome in review, or the company turns out to be one SourceX already knew, which section 3 of the terms treats as a duplicate credited to no later partner. A stall is different. Nothing is decided; the next dated event simply does not arrive. That makes stalls hard to study from one side. From the partner's inbox, a company waiting on its own counsel looks the same as a company waiting on a SourceX step, and both look like a CEO who has lost interest.

The study therefore defines a stall mechanically and leaves the reason for a separate, evidence-bound step.

| Term | Definition used in the study |
|---|---|
| Stage | A span between two dated events, listed in the stage table below |
| Stall | No dated event at the current stage for longer than that stage's threshold, with no ending recorded |
| Ending | A CEO's no, a SourceX outcome, a duplicate ruling, or a recorded payout |
| Restart | A dated event that moves a stalled introduction on |
| Censored | Still open, neither stalled nor ended, on the read date |
| Reason code | The category assigned to a stall from the codebook, backed by a named record |

For keeping a partner's own records straight in the meantime, [the stalled-versus-closed status rules](/guides/the-difference-between-a-stalled-referral-and-a-closed-one) set how a quiet row is labelled; this study uses its own definitions so that both sides' records can be coded the same way.

## Six stages, each with its own clock

Thresholds are study settings chosen so stalls can be counted the same way across partners. They are not SourceX service levels or commitments, and the analysis repeats every count with thresholds half as long and half as long again, to show how much the picture depends on them.

| Stage | Starts with | Ends with | Stall threshold | Whose record dates it |
|---|---|---|---|---|
| 1. CEO asked | The partner's dated question to the CEO | The CEO's answer | 21 days | Partner log |
| 2. CEO agreed | The CEO's written yes | The company's application | 30 days | Partner log, then SourceX |
| 3. Applied | The application | SourceX's verification of it | 30 days | SourceX |
| 4. Verified | Verification | A first conversation between SourceX and the company | 45 days | SourceX |
| 5. In discussion | That conversation | The company decides on scope and terms, or stops | 60 days | SourceX |
| 6. Signed | The company's signed agreement | Acceptance and completed handoff | 90 days | SourceX |

Stage 2 sits inside a published clock of its own: section 2 credits the first valid introduction of a new company within the 90-day attribution window once the company submits a verified application, and the window does not extend. After handoff, the payment stage follows the deadline in section 5, 60 calendar days after the later of the Acquisition Partner's final written acceptance and completed handoff, so it is reported against that published rule rather than a study threshold.

## The stall-reason codebook

Every code needs a named, dated record before a coder may use it. Where no record names a reason, the code is unknown, and unknown is expected to be common.

| Side | Code | Definition | Record required |
|---|---|---|---|
| Company | Deferred to a dated event | The CEO named an event and a date to wait for | The CEO's own message |
| Company | Sponsor unavailable | The person the CEO named to carry it left or was reassigned | A dated note from the company |
| Company | Rights question open | Counsel or a customer contract is being read | The company's statement to SourceX |
| Company | Silent | No reply after the last dated contact | The last dated contact |
| Partner | No owner for the next step | Nobody was named to take the next step | The partner's log shows no named owner |
| Partner | Message misread | The CEO took the note to mean something it did not say | The CEO's words in a dated reply |
| SourceX | Review step pending | The next step sits with SourceX's review | A SourceX stage record |
| SourceX | Program fit pending | Relevance to a current program has not been settled | A SourceX stage record |
| Either | Unknown | No dated record names a reason | None, by definition |

Coders record only the code and the record it rests on, never the private reasons behind a company's choices. The partner-side codes connect to practical fixes elsewhere: [the next-step owner triage](/guides/when-a-referral-has-no-named-next-step-owner) for unowned steps, and [the requested-delay follow-up](/guides/a-referral-follow-up-after-a-company-has-asked-for-more-time) for companies that named their own date.

## Cohort, data sources and consent

- **Cohort:** every introduction submitted from the dashboards of consenting partners during one entry window of two quarters, plus every application SourceX credits to those partners in that window.
- **Sources:** SourceX's dated stage records, and the partners' dated logs of asks and answers, linked by introduction. Only dates, stages and codes are extracted; no company records, documents or figures enter the dataset.
- **Consent:** partners agree in writing before their logs are used. Company-side information is limited to what SourceX already records in the normal course of its review.
- **No exclusions after entry:** duplicates, declines and introductions that never received an answer all stay in, each with its own ending.
- **Read date:** 180 days after the entry window closes.

## Analysis: time in stage, with open cases censored

For each stage, the analysis estimates how long introductions take to reach the next event, treating cases still open on the read date as censored rather than counting them as stalled or as moving. The report then gives, stage by stage, how many introductions entered, moved on, stalled, restarted, ended or remained open, followed by the distribution of reason codes among stalls as counts.

Two coders assign reason codes independently without seeing how the introduction ended; disagreements are reconciled and the agreement rate is reported per side. The report would follow the STROBE checklist for observational studies, whose site describes 22 reporting items, 18 of them shared across cohort, case-control and cross-sectional designs, and notes that it is not a tool for grading study quality (our paraphrase of a search excerpt).

Everything here is descriptive. If introductions from one kind of partner stall less at stage 1, that is an association among these introductions. It does not show that the partner type caused the difference: companies, timing and relationships differ too.

## The table that stays empty until the cohort closes

| Stage | Entered | Moved on | Stalled | Restarted | Ended | Open (censored) |
|---|---|---|---|---|---|---|
| 1. CEO asked | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| 2. CEO agreed | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| 3. Applied | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| 4. Verified | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| 5. In discussion | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |
| 6. Signed | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted | Not yet counted |

Released with the results: the codebook with its version history, both coders' raw codes, the agreement rates, the threshold sensitivity runs and the complete table, zeros included. SourceX-side reasons appear only in aggregate. An annual view of every disposition, of which stalls are one part, is laid out in [the State of Company Data Introductions blueprint](/guides/the-state-of-company-data-introductions-a-research-blueprint).

## Where a stall study misleads

- **Thin partner logs.** A partner who logs only some contacts will show false stalls. Partners whose logs fall below a completeness check are reported separately.
- **Silence gets blamed on the wrong side.** Linking both sides' records reduces this; the unknown code absorbs what linking cannot settle.
- **Thresholds are judgment calls.** The sensitivity runs show how much the counts move when thresholds change.
- **Coders who know endings.** Hiding the outcome field keeps hindsight out of the codes.
- **Consenting partners differ.** Partners who share logs may keep better records and stall less than those who do not.
- **The calendar.** An entry window that spans year-end will show seasonal stalls; the report notes the window's dates.
- **Programs move.** Program criteria can change during the window, and SourceX's screening page says a company that misses one program may suit another later; a change mid-cohort is reported as a dated event.

## 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.

Partners who want their introductions to be measurable later can [register](/register), sign the Channel Partner Agreement before submitting anything, and date every ask and answer from the first message; screen each company against [who qualifies](/who-qualifies) and the [program terms](/terms), and find other study designs in the [guides hub](/guides).

## Sources

- https://refer.sourcex.si/terms
- https://refer.sourcex.si/how-it-works
- https://refer.sourcex.si/who-qualifies
- https://www.strobe-statement.org/checklists/
