Meta Ads Performance Review

Weekend Recovery
& Fragmentation Audit

Magnifiscent UK · act_466067486116405 · GBP · Apr 20, 2026

Period reviewed
Apr 18–20, 2026
Compared against
Apr 10–16 · Mar 20–Apr 9
View Period:
WeekendApr 18 – 20· Daily averages

Purchase ROAS

3.24x

Daily Purchases

16.0

Cost per Purchase

£23.33

Daily Spend

£402.47

CPM

£7.47

Frequency

2.40x

01
Performance Recovery

Yes — significantly.
Above baseline.

The weekend generated 18 purchases per day at a cost per purchase of £27.42 and a Purchase ROAS of 3.51. All three metrics are materially better than both the baseline and the prior 7-day decline period.

Weekend Verdict

Performance is genuinely recovering. However, the improvement is despite the fragmented structure, not because of it.

ROAS & Cost per Purchase — Three Periods

Baseline7-Day PriorWeekend0x0.9x1.8x2.7x3.6x£0£15£30£45£60
ROAS
Cost per Purchase

Account-Level Metrics — Period Comparison (Daily Averages)

MetricBaseline (Mar 20–Apr 9)7-Day Prior (Apr 10–16)Weekend (Apr 18–20)vs Baselinevs 7-Day Prior
Daily Spend£597.73£563.73£493.60-17.4%-12.4%
Daily Purchases13.31118+35.5%+63.6%
Purchase ROAS3.05x2.55x3.51x+15.3%+37.9%
Cost per Purchase£44.99£51.25£27.42-39.0%-46.5%
CPM£13.18£13.65£9.56-27.5%-30.0%
Frequency5.82x5.19x3.21x-44.8%-38.1%
02
Campaign Breakdown

Campaign-Level
Weekend Performance

The most dramatic turnaround is in P2 Scaling Winners — ROAS jumped from 2.23 to 4.22. Retargeting is the standout performer with a ROAS of 5.89 and cost per purchase of just £15.17.

CampaignSpendPurchasesROASCPP
P1 – Creative Testing£759.37243.02x£31.64
Retargeting£151.67105.89x£15.17
Lumamax Testing£208.7852.32x£41.76
P2 Scaling Winners£360.97154.22x£24.06
March Spring Offerspaused

ROAS by Campaign — Baseline vs Weekend

0x2x4x6x8xP1 CreativeRetargetingLumamaxP2 Scaling
Baseline
Weekend

Weekend Spend Distribution

  • P1 Creative Testing
  • P2 Scaling Winners
  • Lumamax Testing
  • Retargeting

Retargeting — Ad Set Detail (Weekend)

Ad SetSpendPurchasesCPPROASFreqLPV→ATCATC→ICIC→Purch
24.1 - Winning Video£110.128£13.777.05x2.1955.1%23.3%80.0%
30.11 - NEW BAN Founder Content£41.642£20.822.82x1.6446.2%25.0%33.3%
03
Fragmentation Analysis

Still fragmented.
Structure unchanged.

The weekend improvement in P2 Scaling is real, but it is being driven by a handful of ad sets that are beginning to exit the learning phase. 24 of 31 ad sets still generated zero purchases while consuming budget.

Meta's recommendation system has issued four separate fragmentation warnings, covering 30, 17, 9, and 5 ad sets respectively. Meta estimates consolidation would deliver a 13% lower cost per purchase.

Meta Warnings Issued

4

Estimated CPP Saving

13%

Account-Wide Waste

£142.79

Top 3 Adsets = Spend

61.2%

P2 Scaling — Ad Set Status24/31 zero-purchase
7 with purchases
24 zero purchases

Total Ad Sets

31

Budget Wasted

£46.55

Waste Rate

17.1%

P2 Scaling Winners — Ad Set Distribution (Weekend)

Ad SetSpend (3 days)PurchasesCost per PurchaseROASStatus
35 Cost Cap£88.115£17.623.42xConverting
5.4 - Transform Home Zenplug£46.321£46.321.27xConverting
7.4 - Many Benefits Zenplug£32.222£16.112.57xConverting
14.4 - iMessage Notification LumaMax£22.921£22.925.93xConverting
14.4 - Us Vs Them Star ZenPlug£17.351£17.359.74xConverting
14.4 - Sold Out 10x LumaMax£10.532£5.2634.56xConverting
14.4 - Myth vs Fact Oil LumaMax£8.291£8.2930.65xConverting
14.4 - Warning Best ZenPlug£6.370Zero Purchases
14.4 - X Signs Scent Upgrade LumaMax£5.980Zero Purchases
14.4 - Copy-Led Expensive LumaMax£5.540Zero Purchases
Remaining 21 ad sets£89.163£29.72Converting
04
Media Buyer Claim Assessment

Five claims.
Assessed against data.

Your media buyer made five specific claims about the account structure and CBO efficiency. Each has been assessed against the live performance data.

Correct1
Partially Correct / Incorrect2
Incorrect2
01
Partially Incorrect

"I don't think the audience etc is a thing of fragmentation — it is set out this way so we know which ad sets are working"

This conflates creative testing with ad set fragmentation. Testing creatives should use multiple ads within a single ad set, not 30+ separate ad sets within a scaling campaign.

02
Incorrect

"The structure is fine"

Meta's own recommendation system issued four separate fragmentation warnings on this account, covering 30, 17, 9, and 5 ad sets respectively.

03
Correct

"Good now"

Performance has improved significantly over the weekend. ROAS of 3.51 and cost per purchase of £27.42 are genuinely strong numbers.

04
Incorrect

"Nothing is wasted as things are battling each other"

Budget waste is quantifiable and significant. £142.79 (9.6% of weekend spend) went to ad sets that generated zero purchases.

05
Partially Correct

"We are being super efficient with budget since it's CBO"

CBO is a budget allocation mechanism, not a fragmentation solution. CBO works best when choosing between a few stable, optimised ad sets — not 31 simultaneously learning ones.

05
Recommended Actions

What to do now.

01
Immediate

Consolidate P2 Scaling

Pause the 24 ad sets in P2 Scaling that received spend over the past 7 days but generated zero purchases. Retain only the proven performers: 35 Cost Cap, 7.4 Many Benefits Zenplug, 14.4 Sold Out 10x, 14.4 Us Vs Them Star, and 14.4 iMessage Notification.

Expected Impact

Estimated 13% CPP reduction per Meta's own recommendation

02
This Week

Freeze New Ad Set Launches

Do not launch new ad sets in P2 Scaling for at least 7 days. Each wave of new ad sets triggers a learning phase that degrades the entire campaign's efficiency for 7–10 days. The Apr 14 batch is still exiting learning.

Expected Impact

Prevents recurrence of the Apr 10–16 performance decline

03
Ongoing

Route New Creatives via P1

Add new creative as additional ads within existing P1 Creative Testing ad sets, not as new ad sets. P1 is correctly structured for testing and is now the account's primary volume driver at £253/day spend and 8 purchases/day.

Expected Impact

Maintains creative testing without triggering fragmentation