Every level of the VSL funnels for the 28 campaigns you listed, joined across Meta Ads, ClickFlare, vturb and BuyGoods. Meta, ClickFlare and vturb are per campaign per day; BuyGoods is the two-day SubID report. vturb views carry the ClickFlare click id, which is how plays and pitch retention are attributed back to each campaign.
A visitor cost $3.33 and was worth $1.96. The gap is not spread across the funnel; it sits in one place.
| Level | What happened | Verdict |
|---|---|---|
| Ad → click | 75k impressions, 4,292 clicks, CPC $3.33, CPM $190 | Expensive, but CTR 5.7% is fine |
| Click → page → play | 100% of clicks tracked, 86% pressed play | No leak |
| Play → pitch (63 min in) | 3,497 plays, 394 still watching at the pitch | 89% gone. This is the lag. |
| Pitch → CTA click | 48% clicked | Fine |
| CTA → order | 188 clicks, 35 orders, 19% | Fine |
Each person who reaches the pitch costs $36 and returns $21. Break-even on this traffic was about 60 orders, which means roughly 670 people at the pitch instead of 394. Nothing before the video and nothing after the pitch has that much headroom.
SodaTide v1 and v2 have the best video and CTA numbers and lose on cost (ROAS 0.73 and 0.61). BFIKL keeps people to the pitch but only 39% click the CTA versus 51 to 58% on SodaTide (ROAS 0.48). SS-MAIN loses them inside the video, 7.3% at the pitch (ROAS 0.25).
Eleven campaigns spent $3,784 for zero orders, most with pitch retention under 8%: the creative brought people who quit the video early. Largest: SSNew118 at $1,063, then ST126, SS135, ST127, SSNew121 and SSNew124 at $360 to $410 each.
ST124 returned 2.13 on $563 with 14% pitch retention. ST118 and SSNew102 broke even. Sunday's SodaTide ran at 0.83 while Sunday's SodaSlim ran at zero on $1,742.
Each lane runs top to bottom. Bars are scaled to the lane's own link clicks. The step rate on the right is against the level above it. Pitch second is 3792s on v1, 3876s on v2. SodaTide spend is split evenly between v1 and v2 because ST campaigns rotate the two fronts 50/50 inside ClickFlare.
BuyGoods only gives the two-day SubID totals, so the daily rows use ClickFlare postbacks for orders. ClickFlare and BuyGoods agree to one order: the extra BuyGoods order arrived with no SubID, so it never posted back.
| Day | Lander | Spend | Impr. | CPM | Link clicks | CPC | VSL visits | Plays | Pitch | Pitch % | CTA | Orders | CR | Revenue | ROAS | CPA |
|---|
CTR was the best of the past two weeks (5.8% Saturday against 4.4 to 6.1% before). The whole CPC jump is CPM: $195 on Saturday and $175 on Sunday against $81 to $155 on every prior day across the same ten accounts. Daily totals below are all campaigns on those accounts, not only the 28.
| Day | Spend | CPM | CTR | CPC | Note |
|---|---|---|---|---|---|
| Sat 22 Aug | $7,722 | $112 | 5.03% | $2.22 | last comparable big day |
| Sun 23 Aug | $4,126 | $95 | 4.36% | $2.19 | |
| Tue 25 Aug | $4,012 | $85 | 4.53% | $1.87 | cheapest day of the period |
| Sat 29 Aug | $1,546 | $113 | 5.39% | $2.09 | |
| Sun 30 Aug | $2,016 | $130 | 5.22% | $2.49 | |
| Wed 2 Sep | $2,828 | $125 | 4.75% | $2.64 | |
| Fri 4 Sep | $3,023 | $155 | 6.09% | $2.54 | 30 campaigns launched |
| Sat 5 Sep | $9,969 | $195 | 5.80% | $3.37 | day 1 for 30 of the 42 campaigns |
| Sun 6 Sep | $4,339 | $175 | 5.40% | $3.25 | 17 of those still running |
30 of the 42 FB campaigns went live on 4 September, each with a $1,000 daily budget, lowest-cost bidding, sales objective, and the identical audience (women 40 to 65, Advantage+). Saturday was their first full day: all in learning, all with a $1,000 budget to fill, all competing for the same impressions. Frequency was 1.03, so it was not the same people seeing the ads twice; it was the auction price for the next impression. On 22 August the same accounts spent $7,722 at a $112 CPM, so volume alone does not explain it. The simultaneous launch does.
Checked against each account's own history: all six NatureArticle accounts ran $104 to $115 CPM from 22 Aug to 4 Sep and $56 to $76 in early August. They are equally healthy and they all jumped together on Saturday. NatureArticle 045's $4.64 CPC is its three creatives' CTR (4.7% against 5.5 to 6.5% elsewhere), not the account. Alphazyl-N-134's $136 was on three campaigns that spent $441 in total and barely pushed; it is also the newest account (created 13 Aug), so account age cuts the other way. Creative-to-creative CPM inside one account spans $150 to $233, the same range as account-to-account. The earlier "account history matters" note was wrong.
Same accounts: $56 to $76 CPM in the first three weeks of August, $104 to $115 from 22 August to 4 September, $177 to $219 on the weekend. The launch spike sits on top of a climb that was already under way, so do not expect Saturday's number to fall all the way back to $110 just by staggering launches.
Break-even at 0.82% visitor-to-order is a $1.96 CPC. The prior two weeks ran $1.87 to $2.83. So even the normal CPC needs pitch retention to move as well. Both levers, not one.
Stagger launches (5 to 8 campaigns a day, not 30). Fewer campaigns with more budget each beats 30 at $1,000. Give new campaigns a cost cap near $2.20 for the first two days instead of lowest cost. Prefer the accounts with cheaper history for new creatives.
Pitch retention per creative with a 95% confidence range. With 100 to 400 plays per creative the ranges are wide, so the verdict column is deliberately cautious: only ST118 is confidently above the 11.3% blend, and no creative is confidently below it yet. Judge a creative on pitch retention after about 300 plays, not 50.
| Creative | Spend | CPC | Plays | At pitch | Pitch % | 95% range | CTA | Orders | ROAS | Verdict |
|---|
Instagram in-app: 690 visits, 13.1% at the pitch, 7 orders (1.0% of visits). Facebook in-app: 3,542 visits, 11.6% at the pitch, 26 orders (0.73%). Same creatives, same ad sets. Worth a placement-split test before assuming it holds.
Clicks between 06:00 and 08:00 ET held 12.6 to 14% to the pitch and produced 12 of the 34 attributable orders. 12:00 and 15:00 to 18:00 clicks held 6.7 to 9%. Almost nothing runs after 19:00 ET, so evenings are untested rather than bad.
All 129 ad sets: women 40 to 65, Advantage+ audience, feed and reels, Facebook plus Instagram, no interests, no custom audiences. Every difference between campaigns is the creative and Meta's delivery draw, not the targeting.
ST116's four ad sets ran 4.8%, 8.8%, 9.3% and 16.9%; ST118's ran 14.6% to 20.3%. That spread on 40 to 80 plays each is sampling noise. It is why the ranking above needs 300 plays before a kill decision, and why the "kill under 8% after 50 plays" rule from the first pass was too aggressive.
4,026 of 4,379 tracked visits are phones (11.8% at pitch, 32 orders). Tablets 260 visits, 12.4%, 2 orders. Desktop 93 visits, 26 plays, no orders. Ad copy angle: "Simple Method" carried 16 orders on $7.3k, "Explore / Learn More" 7 on $3.8k, "Kitchen Recipe" 4 on $1.9k.
The creative-quality view. Pitch % is the share of plays still watching at the pitch second, attributed to the campaign through the click id. ST campaigns rotate v1 and v2, so their vturb numbers are the two players combined. Sorted by spend.
| Campaign | Spend | Link clicks | CPC | VSL visits | Plays | Pitch | Pitch % | $ / pitch | CTA | CTA / pitch | Orders | Revenue | ROAS |
|---|
| SubID | Page | Visitors | Checkout | Visitor→checkout | Orders | Checkout→order | Visitor→order | Commission |
|---|---|---|---|---|---|---|---|---|
| sdtide1 | sodatidesytem.com/sodatide/ | 1,146 | 71 | 6.2% | 12 | 16.9% | 1.05% | $2,880 |
| sdtidev2 | sodatidesytem.com/sodatide-v2/ | 1,166 | 76 | 6.5% | 10 | 13.2% | 0.86% | $2,400 |
| bfikl | natureanditswellness.com/BFIKL/ | 1,577 | 62 | 3.9% | 11 | 17.7% | 0.70% | $2,640 |
| sds1 | wellnessmagicnow.com/v2/ | 357 | 10 | 2.8% | 1 | 10.0% | 0.28% | $240 |
| sds1-bk | SodaSlim backtrap | 21 | 0 | 0% | 0 | — | — | $0 |
| bk1sdt | SodaTide backtrap /bp-sdt1/ | 1 | 1 | — | 0 | — | — | $0 |
| (no SubID) | untagged | 6 | 4 | — | 1 | — | — | $240 |
| v7 · sdtidev4 · v10 | stray tags | 4 | 0 | — | 0 | — | — | $0 |
| Total | 4,277 | 224 | 5.2% | 35 | 15.6% | 0.82% | $8,400 |
Every order in the two-day window (Eastern order time), joined back to its ClickFlare click and its vturb view. "Item" is the last product on the customer record, so an upsell name means the front-end order was followed by that upgrade. Commission is a flat $240 per front-end order; upsells pay $0.
| Order (ET) | SubID | Lander | Campaign | Click day | Reached pitch | Item on record |
|---|
| Front-end SKU | Visitors | Checkout→order | Orders | Commission |
|---|---|---|---|---|
| Sodatide 6 bottles | 2,313 | 16.0% | 20 | $4,800 |
| Sodatide 3 bottles | 2,304 | 13.3% | 2 | $480 |
| Sodatide 2 bottles | 2,304 | 5.9% | 1 | $240 |
| SodaSlim 6 bottles | 1,953 | 17.0% | 10 | $2,400 |
| SodaSlim 3 bottles | 1,951 | 11.8% | 2 | $480 |
| SodaSlim 2 bottles | 1,949 | 0% | 0 | $0 |
Link clicks 4,292 → ClickFlare VSL visits 4,377 → BuyGoods visitors 4,277. Each hand-off is within 3%. Nothing is leaking between the ad and the pixel.
3,497 plays, 394 people still watching 63 to 65 minutes later. After that the funnel is healthy: 48% of the pitch audience clicks the CTA, and 15.6% of checkout visitors buy. Each person who reaches the pitch costs $36 and is worth $21 in revenue. The whole economics of these funnels sits in the first hour of the video.
7.3% pitch retention versus 11.4 to 12.5% on the other three players. Its play rate (87%) and CTA rate are normal. The video itself, not the page or the offer, is why SS-MAIN runs at 0.25 ROAS.
ST124 keeps 14% to the pitch and made 5 orders on $563. SSNew123, SSNew114 and SSNew115 keep 4 to 6% and made none on $547 combined. Same landers, same video, same targeting: the creative is bringing a different audience. The confidence ranges in the creative ranking say the low three are probably weak, not certainly weak.
$1,742 spend, 587 VSL visits, 480 plays, 50 people at the pitch, 15 CTA clicks, zero orders. Saturday the same players took 119 people to the pitch, 51 clicked, 12 bought. Sunday's video numbers were normal; the buyers just weren't there.
2,312 SodaTide visitors, 1 hit on bk1sdt in two days. Either the back-button hook isn't triggering in the Facebook in-app browser or the pixel on /bp-sdt1/ isn't reporting. Worth a device test before trusting the backtrap for revenue.
ST104 and ST125 on Sunday: 339 visits landed on wellnessherenow.net against about $15 of spend. That is bot or review traffic hitting the cloak, not buyers, and it is excluded from every rate above.
SodaTide v1 converts 1.05% of visitors, v2 0.86%, on 12 and 10 orders. Sunday flipped the order (7 vs 2). Not enough orders to call it.
All 33 buyers with a vturb row reached the pitch and clicked the player's CTA; 36 of 37 orders join to a ClickFlare click, and every one of those posted back. The two exceptions are orders placed just after midnight Eastern on Saturday from Friday clicks, and one order with no SubID at all.
$240 whether they take 2, 3 or 6 bottles, paid on every order, and $0 on the 13 upsells. 87% pick 6 bottles. The 12-bottle and Sleep Protocol upgrades are pure network margin.
188 CTA clicks in vturb against 224 checkout-page visitors in BuyGoods. The gap is the "(no SubID)" and stray-tag rows plus people who reload the checkout. The vturb button is the checkout hand-off.
trackingField3 = FB campaign id.conversionTimestamp: visit, so an order counts on the day of the click that caused it. BuyGoods SubID report is Eastern and counts by order time.output/funnel-sep5-6/ in the fb-creative-tracker repo.