Two people, not forty.
But the wiring that did it is still live.

1The answer: 2 of 550, not 40

Every person who submitted a phone number on the two Black Seed Oil pages between Jul 15 and Aug 31 was traced individually. 560 distinct numbers, 550 resolved to a profile, each one's full text history read.

2How it happened

One form, hardcoded to one list, embedded on pages for three different products.

The list binding and the $source label live on the form's submit button, not on the page. One form serving three products means three audiences arrive in one list under one label, and anything sent to that list reaches all of them.

Every campaign that included list SWAHz2 “BFMD — Florida SMS (Jul 2026)”

3The two journeys

Both opted in on a Black Seed Oil page on Jul 15. Numbers masked.

+1612••••042 · opted in 2026-07-15 · 6 BFMD NAD+ texts followed

+1214••••410 · opted in 2026-07-15 · 6 BFMD NAD+ texts followed

Six BFMD NAD+ texts in five days, before a single word about the product they actually asked about. For the second person, the next thing after those six was a Protein Coffee text more than two months later.

4Do the opt-in lists actually deliver?

You asked whether people entering these lists are really opting in and really receiving texts. The three lists behave completely differently, and the best-served group is the one on the “wrong” list.

5How opt-ins are counted, and the cross-check

Two independent sources, and they do not agree. Form submissions are counted from the Submitted Form metric filtered on page_url. List joins are counted from joined_group_at on the list itself.

The cause is double opt-in. List SWAHz2 is set to double_opt_in, so a submission only becomes membership after the person clicks a confirmation. The rest sit in limbo: counted as a form submission, never reachable, never texted.

Two caveats the table shows honestly. Some days have joins with no form activity, which is a bulk import or another source rather than the page, so a single day's percentage is noisy. And the coffee page cannot appear in this table at all, because its hand-coded form never fires the Submitted Form event. For that page only list joins exist.

6Why Protein Coffee's numbers look so much bigger

Three reasons, and only one of them is about the campaign.

Compared properly, on people who actually ended up on a list with consent rather than on raw form submissions, the gap is real but smaller than it looks:

So coffee is running at roughly 1.8x the best Black Seed Oil day on a like-for-like basis, not the 10x that raw submission counts suggest. Most of that advantage is the opt-in setting, not the copy.

7What to change

1. Stop filing one product's sign-ups into another product's list. The exposure here was two people, but only because of when the BFMD sends happened to fall. SWAHz2 still holds Black Seed Oil opt-ins and is still a BFMD audience, so the next overlap is a calendar coincidence away.
2. Wire something to RunR7E. 237 coffee sign-ups today and no flow. The launch is Thursday.
3. Decide single or double opt-in deliberately. Double opt-in is costing roughly two in three sign-ups on the Black Seed list. If it is there for compliance, say so and plan around it. If it is an accident, change it.
4. Give every launch its own form. The list binding and the source label live on the form's submit button, so a shared form cannot keep two audiences apart.
5. Count joins, not submissions. A submission on a double opt-in list is not a subscriber. Reporting submissions overstates the Black Seed Oil result by nearly three times.