Meta US — Sale Period Analysis

Key findings

Two distinct effects are driving the 2026 sale-period CPA increase, and they concentrate in different weeks.

Week after the sale — a creative and audience-quality effect

CTR rose +23.7% YoY while CPC stayed flat (+0.4%) — cheaper, higher-volume delivery. But conversion rate collapsed ‑51.1% and CPA more than doubled (+105%, $116→$239). This is a cheap-clicks, low-intent-traffic pattern, and it lines up with the Jul 1 "Summer NH" creative launch.
The Advantage+ broad structure is structurally weaker than 2025's objective-split campaigns: the 2026 broad Advantage+ campaign during the sale ran at $328 cost/purchase and 0.77% CTR, materially worse than the comparable 2025 "Conversions" objective campaign ($176 cost/purchase, 2.6% CTR).

Sale period itself — a cost/auction effect, not an intent problem

Conversion rate improved during the sale — up +83.8% YoY (1.29%→2.38%). Broad targeting pulling in low-intent traffic would show conversion rate falling, not nearly doubling.
The sale-period CPA increase (+72.9%) is driven by cost, not intent. CPC roughly tripled (+217.9%) and CPM rose 43%, while click volume fell 56% — an auction/cost-of-delivery effect, not a low-intent-audience effect.
The week prior to the sale shows the same cost signature (CPC +81.5%, CTR ‑20.9%, but conversion rate up +40.1%) — this cost pattern predates the sale and isn't unique to Advantage+ or the sale creative.

Period-by-period metrics

All figures aggregated across every active US Meta campaign in the account for each window. Cost per purchase and conversion rate are the two to watch — they move in opposite directions in the sale period vs. the week after.

Visual trend across the six windows

Cost per purchase ($)

Conversion rate (purchases / clicks, %)

CPC ($) vs CTR (%)

Clicks volume

What actually changed in campaign structure

This is the structural context behind the numbers above — worth including in the CEO narrative since it explains why like-for-like comparisons are messy.

2025 — Amazon Prime sale

BAU (before/after) ran as 6 campaigns split by region (Northern / Southern) × objective (Traffic / Conversions / Remarketing).
During the sale itself, BAU paused and 3 consolidated, objective-based campaigns took over: Conversions, Traffic, Remarketing — no region split, but still objective-split.
Remarketing carried the sale: 66% of sale spend, 88 of 100 purchases, $42 cost/purchase.

2026 — Flash Sale

BAU (before/after) runs as Advantage+ campaigns split Northern / Southern, plus separate Traffic and Dynamic Remarketing lines.
During the sale, structure consolidates further into one single broad Advantage+ campaign (no region split) prospecting, plus one combined Remarketing + Retention campaign, plus Dynamic Remarketing.
Broad Advantage+ is the weak link: 50% of sale spend, only 12 of 81 purchases, $328 cost/purchase — Remarketing + Retention and Dynamic Remarketing are both far more efficient ($88 and $32 cost/purchase respectively).

Campaign-level breakdown — sale period

This is where the Advantage+ inefficiency actually lives. Prospecting (Advantage+ broad) is carrying half the sale budget for a sixth of the purchases.

Campaign	Spend	Impr.	Clicks	Purchases	CPC	CTR	Cost/Purchase	Conv. Rate

Recommended next steps

Updated with the ad-level evidence from the Ad performance tab — the sale-period natural experiment (identical creative, broad vs remarketing) and the after-period hook/hold data change what's worth testing versus fixing outright.

Shift budget out of broad Advantage+ prospecting — the sale period proved this directly: identical creative ran at $328.58 CPA broad vs $87.94 in Remarketing + Retention, a 3.7x gap with no meaningful difference in hook rate. Split the broad campaign into themed/segmented ad sets or a narrower lookalike audience rather than full-broad, and re-measure against the Remarketing benchmark.
Fix the "@ashmarieee — Friends Outing" creative before scaling it further — it posts the best hook rate and CTR anywhere in the three-period dataset (45.0% hook, 2.97% CTR) but zero purchases on $566 spend in Northern. Tighten the offer/CTA in the back half of the video or route it to a narrower audience rather than continuing to fund it in broad Advantage+.
Refresh the UGC Yarra Valley creative — its CPA has drifted from $144.84 (before) to $220.64 (after) while running unchanged across all three periods, a fatigue signature rather than a targeting one.
Protect and scale Dynamic Remarketing — $31.63 CPA, the best in the entire dataset, on a product feed with no message-level creative to maintain. Low-effort, high-return relative to everything else in scope.
Retire "Keara — Gift Giver" — the one sale creative with zero purchases in both broad and remarketing placements, suggesting a genuine creative weakness rather than a targeting casualty.

Creative diagnosis, three periods

Ad-level Meta data (Windsor.ai) plus real Hook Rate (video views ÷ impressions) and Hold Rate (ThruPlay ÷ video views) for every video ad, matched by ad name and campaign. CPA target: $58.00.

The cleanest test in the dataset — sale period

The identical six Flash Sale ad concepts ran simultaneously in Advantage+ broad prospecting and in Remarketing + Retention. Same copy, same hooks (16–22% in both placements) — but CPA was $328.58 broad vs $87.94 remarketing, a 3.7x gap. When creative is held constant and only the audience changes, CPA still swings 3.7x. That's a targeting signature, not a creative one.

Where creative genuinely is the problem — week after the sale

The Summer NH flagship video, "@ashmarieee — Friends Outing," posts the highest hook rate of any ad reviewed (45.0% Northern / 39.4% Southern) and the best CTR (2.97%) — but converts to zero purchases on $566 spend in Northern, and $284 CPA in Southern. Broader "group hangout" messaging replaced Spring NH's specific seasonal copy — it pulls attention without pulling intent.

Every ad, before / during / after

Grouped by campaign and placement. Badges flag the headline diagnosis; the line under each card is the specific read for that ad.

Before — Spring NH, Mar BAU (Jun 16–22)

Northern

$1,050 spend · 6 purchases · $175.05 CPA

Couple date (image, testimonial)

under target $66

$66 spend · 1 purchase

Testimonial-driven copy converting efficiently at small scale — continue, consider modest budget increase.

Picnic food (image, 5-star reviews)

biggest spend $743over target $248

42,054 impr · 2.67% CTR · 3 purchases

Reach is outpacing closing power — highest spend in the group but weakest CPA; the volume isn't converting proportionally.

UGC Newport RI (video, partner collab)

hook 42.9%hold 28.8%

$22 spend · 0 purchases

Best hook and hold rates of the entire before-period, on almost no budget — the strongest under-tested candidate to scale.

Spring NH static — GPS won't help

under target $94

4,515 impr · 2.28% CTR · 1 purchase

Solid, unremarkable mid-tier performer — continue as-is.

Spring NH static — made for sharing

under target $15, best CPA in set

$15 spend · 1 purchase

Cheapest purchase in the whole before-period — worth testing at meaningfully higher spend to see if it holds.

Spring NH video — @shroomi.co

hook 35.9%hold 26.6%

$26 spend · 0 purchases

Good engagement, too little spend to judge on conversion yet — fund it further before deciding.

Spring NH video — @explorewithvivi

hold 16.7%, weakest of the videos

$36 spend · 0 purchases

Viewers drop off earlier here than on the other UGC creators — possible message/pacing issue worth a quick creative review.

Spring NH static — treasure hunt

$47 spend · 0 purchases, no video data

No conversions on a real amount of spend — monitor before continuing at this level.

Southern

$2,136 spend · 14 purchases · $152.60 CPA

UGC Yarra Valley (video, testimonial)

biggest spend $1,883over target $145

hook 23.3% · hold 13.3% · 13 purchases

Carries the whole Southern group — 93% of its purchases. Over target, but the only real volume engine in the account at this point in the quarter.

Spring NH static — treasure hunt

$23 spend · 0 purchases

No signal yet on modest spend.

Spring NH static — clues blooming

$57 spend · 0 purchases

Largest zero-conversion static in this group — worth a pause-and-review if it doesn't turn over the next week.

Spring NH static — made for sharing

$27 spend · 0 purchases

No signal yet.

Spring NH graphics — treasure hunt

$25 spend · 0 purchases

No signal yet.

Spring NH graphics — GPS won't help

under target $15

$15 spend · 1 purchase

Efficient on tiny spend, same pattern as its Northern twin.

Spring NH video — @eatinwithlyds

hook 24.0%hold 18.1%

$40 spend · 0 purchases

Healthy engagement, underfunded relative to how well it holds attention.

Spring NH video — @sa_modmama

hook 28.8%hold 19.5%

$31 spend · 0 purchases

Same story — good hook/hold, needs more spend before judging conversion.

3 more test rows

Graphics-sharing $15, Trend Swan Studio $16, ST Swan Studio $5

Negligible spend across all three — no usable signal.

Before — key findings & recommendations

Key findings

Not a low-intent-traffic problem — CTR (1.5–2.5%) and video hook/hold rates (24–43% / 17–29%) are healthy across both regions.
One ad each side is doing almost all the work: UGC Yarra Valley (Southern, 13 of 14 purchases) and Picnic food (Northern, 3 of 6 purchases).
The best-performing UGC creator videos (@shroomi.co, @eatinwithlyds, @sa_modmama, Newport RI) are all under $40 spend — genuinely good engagement that's never been given the budget to prove out on conversion.

Recommendations

Continue Yarra Valley and Picnic food as-is — they're over target but the only real volume engines.
Consolidate spend into the 3–4 best-hook UGC creator videos rather than spreading thin across nine near-identical variants.
Pause static variants sitting at $0 purchases past $40+ spend (clues blooming, Southern) unless they turn over within a week.

During — Flash Sale (Jun 23–30)

Advantage+ prospecting (broad)

$3,943 spend · 12 purchases · $328.58 CPA

TREND_POV (video)

hook 17.6%

$65 spend · 0 purchases

Same creative converts fine in remarketing (see below) — broad placement is the problem, not the ad.

Image with graphics (Food)

over target $607

0.52% CTR · 1 purchase on $607 spend

Weakest CTR of the static sale creative — pause in this placement.

@shaytheordinary — Deal Seeker (video)

biggest spend $2,631over target $263

hook 16.7% · hold 18.5% · 10 purchases

Same ad in remarketing runs at $87 CPA — a 3x gap on identical creative confirms this is a targeting problem.

Image only (Food)

$317 spend · 0 purchases, 0.72% CTR

No conversions on real spend — pause.

Sarzfantasia — Returning Customer (video)

over target $138hold 10.2%, weak

hook 20.3% · 1 purchase

Weakest hold rate of the sale set — this one, unlike the others, may have a genuine creative-fatigue issue on top of the targeting problem.

Keara — Gift Giver (video)

hold 12.0%, weak

$185 spend · 0 purchases

Weak hold and zero purchases in broad — check remarketing performance before writing off the creative entirely.

Remarketing + Retention (same creative, warm audience)

$3,166 spend · 36 purchases · $87.94 CPA

Sarzfantasia — Returning Customer (video)

biggest spend $2,287over target $85

hook 21.2% · hold 15.8% · 27 purchases

Carries the sale — 75% of this group's purchases. Still over target, but the clearest proof that warm-audience placement fixes most of the gap.

@shaytheordinary — Deal Seeker (video)

over target $87

hook 20.0% · hold 22.0% · 5 purchases

3x cheaper than the identical ad running in broad prospecting — same creative, different audience, very different result.

TREND_POV (video)

under target $83

hook 22.0% · hold 25.1% · 2 purchases

Best hold rate of the sale creative set — the message is landing well here, just needs more volume.

Image with graphics (Food)

over target $156

1 purchase on $156 spend

Weakest static performer even in the warm audience — consider pausing or refreshing the visual.

Image only (Food)

under target $31

1 purchase

Efficient on modest spend — continue.

Keara — Gift Giver (video)

$90 spend · 0 purchases

No conversions here either — this is the one sale creative that may be underperforming on its own merits, not just placement.

Dynamic Remarketing

$791 spend · 25 purchases · $31.63 CPA

Dynamic product ad (DPA)

under target $32, best CPA anywhere in the dataset

$791 spend · 25 purchases · 1.38% CTR

No static creative message to critique — it's a product feed. The result underlines that audience warmth, not creative, is doing the heavy lifting here.

During — key findings & recommendations

Key findings

The identical six ad concepts ran in both broad Advantage+ and Remarketing + Retention — CPA was $328.58 broad vs $87.94 remarketing, a 3.7x gap on the same creative. This is a targeting problem, not a creative one.
Hook rates barely move between placements (16–22% both sides) — the creative is drawing attention at roughly the same rate regardless of audience; only the close-through rate changes.
Dynamic Remarketing (product feed, no message-level creative) posts the best CPA anywhere in the entire three-period dataset ($31.63) — audience warmth alone gets most of the way there.
Keara — Gift Giver is the one exception: zero purchases in both broad and remarketing placements, suggesting it may be a genuinely weaker creative rather than a targeting casualty.

Recommendations

Pull spend out of broad Advantage+ prospecting for this creative — test it against a narrower, warmer-adjacent lookalike instead of full-broad.
Protect and scale Remarketing + Retention and Dynamic Remarketing budgets — both are efficient and proven.
Retire or refresh Keara — Gift Giver specifically; it's the only sale creative with no purchases in any placement.

After — Summer NH (Jul 1–7)

Northern

$1,748 spend · 8 purchases · $218.54 CPA

Picnic food (image, 5-star reviews)

biggest spend $882over target $126

2.43% CTR · 7 purchases

The only ad in this group carrying real purchase volume — continue despite missing target.

@ashmarieee — Friends Outing (video)

hook 45.0%, best of all 3 periods0 purchases on $566

2.97% CTR (best in set) · hold 18.3%

The clearest creative problem in the whole dataset — best hook and CTR anywhere, but zero purchases. Broad "group hangout" messaging is pulling attention without buying intent.

Couple date (image, testimonial)

$32 spend · 0 purchases

No signal yet on modest spend.

Summer NH carousel — reviews

$56 spend · 0 purchases

No conversions — pause if unchanged next week.

Test — @a.mom.creates, Family MP (video)

hook 23.0%

$53 spend · 0 purchases

Mid-pack hook, no conversions yet — needs more runway.

Test — Swan Studios, Couple (video)

$53 spend · 0 purchases

No signal.

Test — Swan Studios, VO (video)

under target $53

1 purchase

The only Summer NH test creative with a conversion so far — keep funding to confirm.

Test — Debbie, Couple (video)

$53 spend · 0 purchases

No signal yet.

2 more test rows

Blue skies graphics, Food graphics — under $1 spend each

Negligible spend, no usable signal.

Southern

$2,991 spend · 10 purchases · $299.12 CPA

@ashmarieee — Friends Outing (video)

biggest spend $1,988hook 39.4%over target $284

hold 17.5% · 7 purchases

Same pattern as Northern but with some conversion this time — still the weakest CPA-per-dollar of any high-spend ad in the after-period.

UGC Yarra Valley (video, testimonial)

over target $221

hook 22.9% · hold 13.9% · 3 purchases

Its CPA has drifted up from $145 (before period) to $221 here — likely fatigue on a creative that's been running unchanged across all three periods.

Summer NH graphics — Blue skies

$96 spend · 0 purchases

No conversions on meaningful spend — pause.

Test — Debbie, Couple (video)

$51 spend · 0 purchases

No signal yet.

Summer NH carousel — reviews

$33 spend · 0 purchases

No signal yet.

Test — @a.mom.creates, Family MP (video)

$47 spend · 0 purchases

No signal yet.

Test — Swan Studios, Couple (video)

$46 spend · 0 purchases

No signal yet.

2 more test rows

Food graphics $17, Image only $1

Negligible spend.

After — key findings & recommendations

Key findings

@ashmarieee — Friends Outing is the clearest creative problem in the whole dataset: best hook rate anywhere (45.0% Northern / 39.4% Southern) and best CTR (2.97%), but zero purchases on $566 spend in Northern and $284 CPA in Southern.
Broader "group hangout" messaging (Summer NH) is pulling more attention than Spring NH's specific seasonal copy did, but converting it worse — a genuine creative-driven quality drop, not just a targeting issue.
UGC Yarra Valley's CPA has drifted from $145 (before) to $221 (after) while running unchanged across all three periods — likely creative fatigue on top of the broader-period effect.
Every other Summer NH test creative (Debbie, Swan Studios, @a.mom.creates) has zero purchases after a full week — only "Swan Studios VO" (Northern) has converted at all.

Recommendations

Fix, don't scale, Friends Outing — tighten the message in the back half of the video or route it to a narrower audience before increasing budget further.
Refresh or retire UGC Yarra Valley given the clear fatigue trend across three periods.
Give Swan Studios VO more budget to confirm its early signal; pause the other zero-conversion test creatives if unchanged after another week.

CPA by campaign group, all three periods

Dashed line marks the $58 target. Every group misses it — the question is why.

Hook rate vs hold rate — key video ads

Hook rate (bar) is how well the opening moment stops the scroll. Hold rate (line) is whether the message holds attention through to completion. The Summer NH hero ad has the best hook of any ad here but a mid-pack hold rate — high reach, message doesn't close.

Group-level summary table

Period	Campaign / placement	Spend	Impr.	Clicks	CTR	Purchases	CPA	vs $58 target

Cross-period verdict

Most efficient placements — Dynamic Remarketing ($31.63 CPA) and Remarketing + Retention ($87.94 CPA) during the sale, both on identical or near-identical creative to the weakest performer below.
Weakest placements — Sale-period Advantage+ broad ($328.58) and After-period Southern Summer NH ($299.12) — the two furthest from target in the dataset.
Reading — every time the same or similar creative runs in a narrower, warmer placement, CPA improves 3–4x. Creative does matter — the after-period's broad-appeal Summer NH copy visibly pulls a different, less bottom-funnel audience than Spring NH's specific copy did — but the size of that effect is smaller than the size of the targeting effect.

Cumulative CPA vs $58 target, May 1 – Jul 13 2026

Account-level daily CPA (Windsor.ai), mapped against the three creative periods covered in this analysis.

Spring NH (BAU) — through Jun 22 Prime Sale (Flash Sale) — Jun 23–30 Summer NH — from Jul 1

What the trend actually shows

May was the only period consistently near target. CPA opened at $53.71 on May 1 and drifted up to $68.74 by May 31 — degradation was already underway well before either the sale or the Summer NH launch.
Spring NH BAU (June) ran consistently worse than the sale itself. CPA sat in the high $70s–low $80s for nearly all of June ($79–$82 for nine straight days, Jun 6–14) — meaningfully worse than the $75–78 the account posted during the actual sale week (Jun 23–30).
The sale week was a mild relief, not a crisis. CPA eased from $78.29 (Jun 23) down to $75.05 (Jun 30) — Remarketing + Retention and Dynamic Remarketing pulling the blended number down, consistent with the Ad performance tab.
The real inflection point is Jul 1–5, right at the Summer NH launch. CPA cratered to $41.61 on Jul 1 (likely a low-spend/attribution-lag artifact at launch), then climbed in a straight line to $90.96 on Jul 5 — the single worst day in the entire 74-day window — and stayed elevated ($82–$90) through Jul 13.
Reading — this is the strongest chart-level evidence for your creative hypothesis: the worst sustained CPA of the whole quarter starts exactly when Summer NH creative went live, not when the sale started or ended.