Charlene FDAS Email Investigation
Private QC document. Enter hub password to continue.
That's not right. Try again.
The Missing Low-Score Results Email
Verdict
The results email was genuinely never sent. Most likely Sparrow's GoHighLevel automation stalled or errored on this contact; the submission itself almost certainly completed (the confirmation page rendered). If that's right, it's a real bug hitting real low-scoring prospects, the very people who most need the follow-up, and it's worth fixing before the fall-cohort push.
Run 1 vs Run 2 timeline
| Run 1: Charlie Bennett (high score) | Run 2: Charlene Foster (low score) | |
|---|---|---|
| Submitted | 2026-07-05 evening, via the Boulder FDAS page | 2026-07-05 evening, same page, same widget |
| Email used | [email protected] | [email protected] |
| Confirmation page | Shown ("scores are being calculated") | Shown ("scores are being calculated") |
| Results email | Arrived 2026-07-06 02:14 UTC, about 2 hours after submitting | Never arrived (3+ days, re-verified 2026-07-08) |
Evidence
- Run 1's email proves the pipeline works end to end: from Rebuilding Seminars ([email protected], signed Sparrow Gabriel), subject "Your Self-Test Results and Upcoming Seminars," delivered through the same Cloudflare catch-all that Charlene's address uses.
- Gmail was searched exhaustively with
in:anywhere(which covers spam and trash): zero results of any kind to [email protected], no stray confirmation, nothing from rebuildingseminars addressed to her after July 4. - Filtered/missed mail is ruled out: the catch-all forward is verified working (run 1's email arrived through the exact same route the same evening).
- A failed submission is unlikely: the /calculating confirmation page only renders after a successful form POST, and both runs saw it. Only the low-score run's email vanished.
- Residual long shot: a per-address bounce at the catch-all. Possible in principle, but the catch-all is address-agnostic; a GHL-side stall is far more probable.
The exact question for Bob to pass to Sparrow
"A test taker named Charlene Foster ([email protected]) completed the Boulder FDAS survey on the evening of July 5 and saw the confirmation page, but never received the results email (another taker the same evening got theirs in about 2 hours). Can you check in GHL: (a) does her contact and survey submission exist, (b) did the results-email automation fire for her, and if not, what step did it stop at? Her answers were heavily on the struggling end, so please also check whether the automation branches or fails on low scores."
Retest attempt (2026-07-09): blocked by reCAPTCHA
The low-score retest as Charlene Foster ([email protected]) couldn't be submitted automatically. The whole survey fills perfectly, all 100 answers on the struggling end, the same low-adjustment demographics, name and email, and no missing required fields. It's the final step that stops us: the submit slide carries a Google reCAPTCHA "I'm not a robot" checkbox, and the survey's backend rejects any automated submit with "Requires captcha verification."
It was reproduced three ways (plain headless, real Chrome with human-paced typing and mouse movement, and a warmed browser profile). Every run ended the same: the "I'm not a robot" box stays unchecked and the form shows "Please fill captcha to proceed," and the backend returns a 429 captcha error. So no results email to [email protected] will arrive from this attempt, since the submission never actually landed.
That's why run 1 (Charlie Bennett) went through and this one didn't: run 1 was completed through a real, logged-in browser that clears the captcha, which a headless script can't do. The clean way to finish the low-end check is to run the same fill through a real logged-in Chrome and tick the checkbox by hand, then watch the inbox about 2 hours later.
Worth naming plainly: this doesn't change the core finding. The original anomaly (run 1's low-score email never sending after a genuine, successful submit) is still the open question for Sparrow. This retest just couldn't add a second data point because the captcha wall blocks automated submission.
Why this matters beyond QC
If the automation genuinely stalls on low scorers, every real prospect who scores low (the rawest, most-in-need people, and the likeliest to enroll) is silently getting nothing after being told "your scores are being calculated." That's a lead-losing bug in Bob's funnel, worth fixing before the September 20 cohort push.
Next steps
- Finish the low-score retest through a real logged-in browser (the automated attempt was blocked by the survey's "I'm not a robot" checkbox): fill it, tick the captcha by hand, then watch [email protected] at the ~2-hour mark. If the email arrives, the July 5 miss was a one-off; if it doesn't, the low-score stall is reproducible and Sparrow has two concrete cases to trace.
- Send Sparrow the question above (via Bob), now strengthened with the retest data point.
- Close the low-end engine verification: once a low-score results email lands, read the chart values and confirm they plot low (below the Target 80 line). High end is already confirmed correct (run 1 scored perfect 100s).