Spoiler Golf · Rosewater Growth · October 12, 2026

Meta Structure and Performance Overview

Same CPA as the old setup for the first 17 days at 23% more daily spend; since mid-September cost per purchase has run $286 to $389 a week as CPMs rose and new creative took a third of spend.

The structure is not what moved CPA: 45 days before vs 45 days after, $252 to $292

Meta platform data, 7-day click / 1-day engaged view. Before = Jul 9 to Aug 22 on the old Advantage+ campaign. After = Aug 28 to Oct 11 on the consolidated structure (Aug 23 to 27, the switch-over week, is left out of both).

Spend per day

$784

was $649 +21%

Purchases

121

was 116 +4%

Cost per purchase

$292

was $252 +16%

ROAS

1.08

was 1.24 -13%

CPM

$32.42

was $22.08 +47%

Cart to purchase

20.9%

was 16.0% +31%

The first 17 days on the new structure matched the old one. Aug 28 to Sep 13 ran $250 per purchase and 1.24 ROAS on $795 a day, against $252 and 1.24 on $649 a day before the change, while CPMs were already 58% higher.

CPA rose later, inside the same campaign. Sep 14 onward ran $286 to $389 per purchase a week with nothing structural changed. What changed was the market (Q4 auction costs, the October Prime event) and the creative mix (61 new ads launched since Sep 28 took 33% of spend and have not found a winner yet).

Once people reach the site, the new setup converts better. 0.62% of landing page visits became a purchase vs 0.46% before, and one in five carts closes vs one in six. The cost is getting the click: impressions cost 47% more and outbound click rate dipped from 2.06% to 1.85%.

One campaign then, one campaign now: what changed is bid control and creative supply (24 ads to 183)

Before, to Aug 22

Advantage+ shopping campaign, highest volume

1 campaign, 1 ad set, no cost control: Meta spends the budget wherever it can.
24 ads delivered in the 45 days; the top two carried 39% of spend.
Several winners were 7 to 25 months old and declining week over week.
Advantage+ shopping blends new and existing customers by default.
$252 per purchase against a $180 target.

Now, from Aug 23

Consolidated CBO, cost per result goal

1 campaign, 1 ad set holding every proven ad plus each new batch.
A cost per purchase goal of $175 (tested at $150, $180 and $200).
183 ads delivered in 45 days; new batches land next to the winners.
A second campaign on a 1.5 ROAS goal ran alongside for the first weeks; it took $1,851 for 2 purchases and was paused, with its ads folded into the main campaign.
Attribution set explicitly to 7-day click / 1-day engaged view.

Why this shape fits Spoiler at today's size

Meta needs about 50 purchases a week per ad set to learn. At a $180 target that is roughly $1,285 a day for one ad set (target × 7.14). Spoiler spends $650 to $1,000 a day, about 15 to 25 purchases a week, so every split into more campaigns or ad sets would leave each one short of signal. One ad set is the right size; the old setup was one too.

The cost goal caps the downside. When Meta cannot find purchases near the goal, it spends less instead of buying expensive ones. That is why daily spend fell to $628 in Sep 14 to 27 while results softened, instead of running the full $1,000 at a worse cost. Meta treats the goal as an average to aim for rather than a hard ceiling, so reported CPA can still sit above it.

New creative tests inside the proven environment. New ads compete for spend next to the ads that already convert, so a new winner can earn budget the day it proves itself, and nothing is fractured into separate test campaigns.

What we read, and how often

Metric	Why it matters	Where it sits now (Aug 28 to Oct 11)
Cost per purchase	The efficiency target the bid is set against ($180 goal, $175 control)	$292
ROAS (7dc / 1d engaged view)	Revenue Meta can credit per dollar, in one consistent attribution window	1.08
Spend delivered vs budget	Under-delivery is the signal that creative cannot hit the goal	$784 a day vs $1,000
Outbound click rate	Whether the ad earns the click; the step new launches are missing	1.85%
Add to cart / landing page view	Whether the page and offer convince	2.98%
Cart to purchase	Whether checkout closes	20.9%
New creative share of spend	Whether fresh ads are earning budget at all	33% (Sep 28 to Oct 11)

Every read is taken on the last 7, 14 and 30 days; an ad or campaign counts as declining only when all three agree. A new batch gets 7 to 10 days before it is judged, because the attribution window itself is 7 days.

Before vs after: purchases up 4% on 21% more spend, conversion up, impression cost up 47%

Meta, 7dc / 1d EV	Before (Jul 9 to Aug 22)	After (Aug 28 to Oct 11)	Change
Spend	$29,190	$35,294	+21%
Spend per day	$649	$784	+21%
Purchases	116	121	+4%
Revenue (Meta-credited)	$36,244	$37,948	+5%
ROAS	1.24	1.08	-13%
Cost per purchase	$252	$292	+16%
AOV	$312	$314	0%
CPM	$22.08	$32.42	+47%
Outbound CTR	2.06%	1.85%	-10%
Add to cart / LPV	2.89%	2.98%	+3%
Cart to purchase	16.0%	20.9%	+31%
Purchase / landing page view	0.46%	0.62%	+35%
Ads with delivery	24	183

Motion, Meta platform data at the account's 7-day click / 1-day engaged view setting. Green better, red worse, grey within 3% or a spend change.

Inside the new structure: $250 per purchase to Sep 13, $286 to $389 a week since

Window	Spend / day	Purchases	CPA	ROAS	CPM	oCTR	ATC / LPV	Cart to purch.	Ads live
Before (old), Jul 9 to Aug 22	$649	116	$252	1.24	$22.08	2.06%	2.89%	16.0%	24
Aug 28 to Sep 13	$795	54	$250	1.24	$34.88	2.14%	3.70%	18.8%	68
Sep 14 to 27	$628	27	$326	0.99	$35.59	2.14%	2.69%	20.0%	65
Sep 28 to Oct 4	$1,022	25	$286	1.10	$49.24	1.91%	3.35%	25.5%	42
Oct 5 to 11	$833	15	$389	0.79	$18.88	1.22%	1.57%	25.9%	128

Oct 5 to 11 has 15 purchases, so treat it as a signal, not a read. That week the new batch went live (ads with delivery jumped from 42 to 128), impressions got cheaper and click rate fell: Meta was testing the new ads on broad, low-cost inventory.

Cost per purchase by window

Before (old)

$252

Aug 28 to Sep 13

$250

Sep 14 to 27

$326

Sep 28 to Oct 4

$286

Oct 5 to 11

$389

By campaign, Aug 28 to Oct 11

Campaign	Spend	Purchases	CPA	ROAS	CPM	oCTR	AOV	Ads
Consolidated Top Ads, cost per result goal (CC)	$33,444	119	$281	1.12	$31.93	1.84%	$314	158
Consolidated Top Ads, 1.5 ROAS goal (RG), paused	$1,851	2	$925	0.33	$44.92	1.91%	$308	25

The ROAS-goal campaign was the side-by-side bid test; it did not earn its budget and its ads now run only in the main campaign.

Three drivers: impressions cost 47% more, new launches click at 1.14%, proven ads hold

The auction got more expensive. CPM went from $22 in July and August to $32 across the new period and $49 in the week of Sep 28. Late September through Q4 is when the biggest advertisers ramp up, and the October Prime event pulled attention away from DTC the week of Oct 5. Every brand we run saw conversion rates soften that week.

The old setup also bought cheaper reach. Its top ad (OG Stealth, 20% of spend) ran at an $11.88 CPM with a 12% all-click rate but only 1.9% outbound clicks: a lot of low-cost, low-intent inventory. Cheaper reach helped hold the old CPA while spend stayed around $650 a day.

New creative has not found a winner yet. The 61 ads launched since Sep 28 took $4,271 (33% of spend) for 8 purchases at $534 each, with a 1.14% outbound click rate against 1.85% for the account. Once a click lands, they convert like the rest (3.2% add to cart). The gap is the click.

Proven ads perform about as they did before. The same creatives in both structures land on either side of their old numbers, which is what a market shift looks like rather than a structure problem (table below).

Same ads, old structure vs new

Creative	Old: spend	Old: purch.	Old: CPA	New: spend	New: purch.	New: CPA
Stupid Good	$2,410	14	$172	$2,832	17	$167
Best Putter 3	$2,177	6	$363	$3,757	15	$250
OG Stealth Plain White	$5,819	18	$323	$1,730	6	$288
Best. Testimonial, Low High	$2,390	18	$133	$1,350	6	$225
Lead Edge Roll Explainer	$5,681	20	$284	$2,875	8	$359
Roll	$1,545	8	$193	$2,730	10	$273
Best Putter 2025	$924	5	$185	$1,041	3	$347

Old = Jul 9 to Aug 22, new = Aug 28 to Oct 11. Most rows rest on fewer than 15 purchases, so read them as direction only. Three improved, four got more expensive, in line with the 47% rise in impression cost.

Top 10 ads by spend since the switch

Ad	First launched	Spend	Purchases	CPA	ROAS	oCTR	Add to cart
Best Putter 3	Feb 24, 2026	$3,757	15	$250	1.29	2.83%	105
He Already Has Everything (gift static)	Sep 11, 2026	$3,518	14	$251	1.24	2.78%	74
Lead Edge Roll Explainer	Aug 23, 2026 copy	$2,875	8	$359	0.88	2.83%	40
Stupid Good	Oct 13, 2025	$2,832	17	$167	1.86	2.10%	61
Roll	Aug 23, 2026 copy	$2,730	10	$273	1.21	0.56%	32
OG Stealth Plain White	Mar 25, 2026	$1,730	6	$288	1.31	1.75%	32
Best. Testimonial, Low High	Aug 20, 2025	$1,350	6	$225	1.46	1.95%	36
V07, founder-skeptic video	Sep 16, 2026	$1,198	2	$599	0.51	3.71%	4
VIDEO-24, Don't Just Buy a Better Putter	Sep 30, 2026	$1,190	0	n/a	0.00	2.12%	16
Best Putter 2025	Jan 4, 2026	$1,041	3	$347	0.87	1.87%	20

Aug 28 to Oct 11, all in the cost per result goal campaign. The September gift static is the one new ad so far to earn top-two spend at the account's best CPA; the two newest videos win the click (2.1% to 3.7%) but have not yet turned it into carts.

Keep the structure, fix the click on new launches, judge the October batch on Oct 12 to 18

Hold the consolidated campaign and the $175 cost per purchase goal through October; it is what keeps spend from running ahead of results in a soft month.
Judge the 61-ad batch launched since Sep 28 on Oct 12 to 18, its first full 7-day attribution window; ads with two weeks of little spend come out so budget stays on the ones earning it.
Close the click gap on new launches (1.14% vs 1.85%): partnership and creator-led video and the founder origin story, built inside the approved messaging territories in your brand positioning document.
Brief iterations of the three ads closing at $167 to $251 (Stupid Good, Best Putter 3, the gift static): new openings and formats on the same proven message, and monitor whether each earns more budget.
Keep creative volume on the Creative Demand Model through November, with the Black Friday batch submitted by Nov 20; the holiday window is where last year's account ran $93 a purchase.
If new launches have not brought weekly CPA back toward $250 by the first week of November, test a highest-volume twin campaign carrying only the proven ads, judged on account-level CPA after two weeks of stable spend.

All figures are Meta platform data via Motion at the account's 7-day click / 1-day engaged view setting, Jul 9 to Oct 11, 2026.