Ad Quality & Cost Impact Report
Juicy Marbles - Cross-Platform Analysis
Executive Summary
Your ad quality directly affects what you pay per click and how much revenue each dollar generates. This report analyzes quality signals across Meta and Google Ads, quantifies the cost impact, and identifies where quality improvements would save the most money.
Average Quality Score of 4.6 (vs 7.0 in US) is inflating your European CPCs. Generic keywords are scoring QS 1-3 while brand terms score 10.
Google keywords with spend but zero conversions. $4,937 in Meta ads with >$100 spend and zero tracked ROAS.
Top-performing ads (11-14x ROAS) are running at high frequency - still converting well but at risk of audience burnout.
Part 1: Meta Ads Quality Analysis
In-House vs FC Agency Performance
| Metric | In-House | FC Agency | Gap |
|---|---|---|---|
| Total ads | 274 | 490 | FC ran 79% more ads |
| Total spend | $98,996 | $148,761 | FC spent 50% more |
| Total revenue | $570,102 | $529,310 | IH earned 8% more on less spend |
| Weighted ROAS | 5.76x | 3.56x | IH 62% more efficient |
| Average CPC | $0.59 | $0.84 | FC pays 42% more per click |
| Average CTR | 3.02% | 2.50% | IH gets 21% more clicks |
The quality gap costs real money. FC spent $148,761 and generated $529,310 in revenue. At in-house efficiency (5.76x ROAS), that same $148,761 would have generated $856,693 - an additional $327,383 in lost revenue.
Why In-House Ads Perform Better
Your top 15 revenue-generating ads tell the story:
| Rank | Source | Ad | Spend | ROAS | Revenue |
|---|---|---|---|---|---|
| 1 | In-House | Gear Patrol FF | $11,019 | 7.40x | $81,501 |
| 2 | In-House | Fake testimonials FF | $8,296 | 8.37x | $69,447 |
| 3 | In-House | Ribs Pic 2 (may 2024) | $4,194 | 11.63x | $48,778 |
| 4 | In-House | Canon 17 sec IG Story Ad | $3,658 | 10.66x | $39,002 |
| 5 | FC | Sabrina All Benefits Mix 2 | $7,586 | 4.12x | $31,251 |
| 6 | In-House | Fake testimonials FF (2) | $3,355 | 9.02x | $30,267 |
| 7 | In-House | BBQ video 2 (june2025) | $6,026 | 3.28x | $19,743 |
| 8 | In-House | Jan 2025 30% off | $1,093 | 13.86x | $15,144 |
| 9 | FC | Sabrina Baby-ribs curiosity | $2,944 | 4.93x | $14,520 |
| 10 | In-House | Versatility (apr 2024) | $1,459 | 9.77x | $14,258 |
The pattern: in-house ads consistently hit 7-14x ROAS. FC ads cap out around 4-5x. The best FC ad ($31,251 revenue) would rank only 5th among in-house performers.
What makes in-house ads better
- Higher CTR (3.02% vs 2.50%) means the creative resonates more - people click more often
- Lower CPC ($0.59 vs $0.84) is a direct consequence of higher relevance scores
- The brand voice in in-house ads (chef-forward, irreverent) drives curiosity clicks, which FC's more generic approach does not achieve
Money Losers
Sub-1x ROAS ads (>$100 spend, losing money):
| Source | Ad | Spend | ROAS | Loss |
|---|---|---|---|---|
| FC | Sabrina All lead-gen Lifestyle 3 | $574 | 0.52x | -$275 |
| FC | Sabrina Whole-cut-loin curiosity | $464 | 0.48x | -$240 |
| FC | Sabrina Ribs UGC | $263 | 0.25x | -$196 |
| IH | BBQ video 2 variant | $471 | 0.63x | -$172 |
| FC | Sabrina Thic-cut-filet Salesy | $786 | 0.79x | -$165 |
8 of the top 10 money losers are FC agency ads. Total identified loss across sub-1x ads: approximately $1,593 in direct losses.
Zero-ROAS Ads (>$100 spend, no tracked purchases)
| Ad | Spend |
|---|---|
| Waitrose x JM Valentine's feb25 | $1,000 |
| UK - Sainsburys (Static) | $557 |
| Content Creator Test Emily (3 variants) | $1,206 |
| Filet video ad - Meny stores | $349 |
| UK - Sainsburys (Reviews) | $195 |
| UK Video Ad November 2025 | $183 |
These are mostly UK retail and creator test ads - $4,937 total. The UK/retail ads may generate in-store sales not tracked in Meta's pixel, but the creator tests produced nothing measurable.
Fatigued but Still Performing
These ads have high frequency (audience seeing them 3+ times) but are still generating strong returns:
| Ad | Frequency | ROAS | Revenue |
|---|---|---|---|
| Ribs Pic 2 (may 2024) | 3.6 | 11.63x | $48,778 |
| Canon 17 sec IG Story Ad | 3.3 | 10.66x | $39,002 |
| Sabrina All Benefits Mix 2 (FC) | 3.4 | 4.12x | $31,251 |
| Fake testimonials FF | 3.7 | 9.02x | $30,267 |
| Jan 2025 30% off | 5.4 | 13.86x | $15,144 |
| Versatility (apr 2024) | 4.0 | 9.77x | $14,258 |
| Bundles Static Oct 2024 | 4.7 | 7.78x | $10,953 |
"Jan 2025 30% off" at frequency 5.4 and 13.86x ROAS is remarkable - the audience has seen it over 5 times on average and it still converts at the highest ROAS in the account. This suggests the offer is strong enough to overcome fatigue.
However, frequency above 3.5 typically signals the beginning of performance decay. The creative refresh system is designed to catch this transition and prepare replacements before performance drops.
Part 2: Google Ads Quality Score Analysis
The EU Quality Score Problem
Google assigns each keyword a Quality Score from 1-10 based on three components: creative quality, post-click experience, and predicted click-through rate. Higher QS means lower CPCs - Google literally charges you less when your ads are more relevant.
| Region | Average QS | Keywords Scored | Weekly Spend |
|---|---|---|---|
| US | 7.0 | 31 keywords | $518 |
| EU | 4.6 | 50 keywords | $163 |
EU Quality Score Distribution
| QS | Count | Keywords |
|---|---|---|
| 10 | 7 | juicy marbles, juicy marbles vegan meat, juicy marbles steak, juicy marbles filet, etc. |
| 9 | 3 | juicy marbles ribs, etc. |
| 7 | 4 | vegan steak, etc. |
| 6 | 3 | vegan meat alternatives, vegan ribs |
| 5 | 5 | vegan lunch ideas, etc. |
| 4 | 6 | plant based meat, plant based bacon, etc. |
| 3 | 5 | various generic terms |
| 2 | 4 | vegan food, etc. |
| 1 | 13 | vegan meal, vegan protein meals, high protein vegan meals, non meat protein sources, etc. |
13 keywords at QS 1 - the worst possible score. All are generic terms like "vegan meal" and "high protein vegan meals." Brand terms score 9-10.
EU Quality Score Components
| Component | Average (1-4 scale) | Distribution |
|---|---|---|
| Creative Quality | 2.9 | Below avg: 18, Avg: 17, Above avg: 15 |
| Post-Click Quality | 2.9 | Below avg: 21, Avg: 13, Above avg: 16 |
| Predicted CTR | 2.5 | Below avg: 33, Avg: 7, Above avg: 10 |
Predicted CTR is the weakest component - 33 of 50 keywords score "below average." This means your ad copy is not compelling enough for the keywords you are bidding on, or the keyword-to-ad match is poor.
US Quality Score Distribution
| QS | Count | Keywords |
|---|---|---|
| 10 | 8 | juicy marbles, juicy marbles vegan meat, juicy marbles steak, marble vegan steak, etc. |
| 8 | 9 | vegan steak, buy juicy marbles, vegan meat, plant based steak, etc. |
| 7 | 4 | various terms |
| 6 | 3 | vegan ribs, etc. |
| 5 | 2 | vegan meat alternatives, etc. |
| 3 | 2 | generic terms |
| 1 | 2 | high protein vegan meals, etc. |
US is healthier - brand terms at 10, most generic terms at 7-8. Only 4 keywords below QS 5.
Cost Impact of Low Quality Scores
Google's auction system uses QS to adjust what you pay. The relationship is roughly:
- QS 10: you pay ~50% less than baseline
- QS 7: baseline (neutral)
- QS 5: you pay ~25% more
- QS 3: you pay ~67% more
- QS 1: you pay up to ~400% more
EU CPC Premium from Low QS
| Keyword | QS | Spend/wk | CPC | Est. Fair CPC | Premium |
|---|---|---|---|---|---|
| plant based meat | 4 | $3.66 | $0.41 | $0.23 | 78% |
| vegan meat alternatives | 6 | $2.26 | $0.56 | $0.48 | 17% |
| vegan food | 2 | $1.95 | $0.39 | $0.11 | 254% |
| vegan lunch ideas | 5 | $0.82 | $0.21 | $0.15 | 40% |
Total estimated CPC premium waste in EU: $4.38/week ($228/year) on matched keywords.
US CPC Premium
The big one: "high protein vegan meals" at QS 1 spent $61.29 in one week at $1.20 CPC with only 1 conversion. At QS 7, that same keyword would cost roughly $0.17/click. Estimated weekly waste on this keyword alone: $53.
Total estimated US CPC premium waste: $53.38/week ($2,776/year) - almost entirely from a single QS 1 keyword.
Zero-Conversion Keywords
US - $143/week on keywords with spend but zero conversions:
| Keyword | Spend/wk | Clicks | CPC |
|---|---|---|---|
| high protein vegan meals (QS 1) | $18.91 | 15 | $1.26 |
| vegan steak (QS 8) | $7.79 | 13 | $0.60 |
| juicy marbles ribs | $13.28 | 19 | $0.70 |
| buy juicy marbles | $5.61 | 3 | $1.87 |
| + 39 more keywords | $97.49 | - | - |
EU - $31/week on keywords with spend but zero conversions. Lower absolute waste, but the same pattern: generic keywords cost more and convert less.
Part 3: Combined Cost Impact
What Low Quality Costs You Annually
| Issue | Weekly Cost | Annual Cost |
|---|---|---|
| Google US QS premium (mainly "high protein vegan meals" at QS 1) | $53 | $2,776 |
| Google EU QS premium (13 keywords at QS 1) | $4 | $228 |
| Google zero-conversion keywords | $174 | $9,048 |
| Meta FC efficiency gap (extrapolated from historical) | - | ~$327,383 in lost revenue |
| Meta zero-ROAS ads (creator tests, UK retail) | - | $4,937 total |
| Meta sub-1x ROAS ads | - | $1,593 in direct losses |
Priority Actions by Impact
Kill or restructure "high protein vegan meals" in US Google Ads
QS 1, $61/week spend, 1 conversion. Either pause this keyword, improve ad copy relevance, or create a dedicated ad group with tightly matched ad copy.
Retire underperforming FC creative
8 of top 10 money losers are FC ads. Sub-1x ROAS FC ads should be paused and replaced with in-house creative. The ROAS gap (3.56x vs 5.76x) represents the single largest quality issue in the account.
Improve EU keyword-to-ad relevance
13 keywords at QS 1 in EU, mostly generic plant-based/vegan terms. Create dedicated ad groups with ad copy specifically written for these terms. Focus on predicted CTR (weakest component at 2.5/4).
Pause or restructure zero-conversion Google keywords
43 US keywords and 21 EU keywords spending $174/week with zero conversions. Review conversion attribution window. Add negative keywords to prevent irrelevant matches.
Prepare creative refresh for fatigued winners
7 ads at frequency 3.3-5.4 generating $189,654 in revenue. These are your best performers - protect them by having replacements ready before performance drops.
The Quality Score-to-Cost Relationship
For every QS point you gain, your CPC drops approximately 13-16%. Here is what improving your EU average from 4.6 to 7.0 would look like:
| Current EU Avg QS | Target QS | Est. CPC Reduction | Weekly Savings | Annual Savings |
|---|---|---|---|---|
| 4.6 | 5.0 | ~6% | ~$10 | ~$520 |
| 4.6 | 6.0 | ~18% | ~$29 | ~$1,530 |
| 4.6 | 7.0 | ~28% | ~$46 | ~$2,370 |
Part 4: Quality Improvement Roadmap
Google Ads - Immediate (Week 1)
- Pause "high protein vegan meals" in US until ad copy is rewritten to match the keyword intent. Current QS 1 means Google considers your ad nearly irrelevant to this search.
- Review all QS 1-3 keywords in EU (22 keywords). For each, decide: improve ad relevance, add as negative keyword, or pause.
- Check conversion tracking - many high-traffic keywords show 0 conversions in 7 days. Verify your conversion window and attribution model.
Google Ads - Short Term (Weeks 2-4)
- Single Keyword Ad Groups (SKAGs) for your top 10 spending generic keywords. Write ad copy specifically matching each keyword's intent.
- Landing page optimization - Post-Click Quality averaging 2.9/4 in EU means your landing pages need work for these keyword audiences.
- Expand brand keywords - Brand terms score QS 9-10 and convert well. Capture more brand + product combinations.
Meta Ads - Immediate
- Audit all active FC ads - Pause anything below 2x ROAS that has been running 30+ days.
- Scale in-house winners - Your top in-house ads are generating 7-14x ROAS. Ensure budget allocation favors these.
- Creator test protocol - Three "Content Creator Test Emily" variants spent $1,206 with zero ROAS. Set a kill threshold (e.g., $200 spend with 0 conversions = pause).
Meta Ads - Ongoing
- Creative refresh cycle - The automated system detects frequency fatigue and generates replacement copy. Use it to protect your top performers.
- Copy QA system - Every new variant goes through the 5-gate quality check before reaching your review queue. This prevents the quality drift that characterized FC creative.
Part 5: Landing Page & Page Speed Analysis
Why This Matters
Google's Quality Score has three components: Ad Relevance, Expected CTR, and Landing Page Experience. Your EU account's Post-Click Quality averages 2.9 out of 4 - meaning Google considers your landing pages below average for the keywords you're bidding on. This directly inflates your CPCs.
Landing Page Traffic Concentration
US (48 unique URLs, last 7 days)
| Landing Page | Clicks | Impressions | Cost | CTR |
|---|---|---|---|---|
| juicymarbles.com (homepage) | 783 | 32,822 | $376 | 2.4% |
| /collections/all-products | 53 | 4,228 | $58 | 1.3% |
| Shopping product pages (various) | 37-54 each | 500-2,800 | $43-50 each | 2-7% |
| /blogs/recipes | 15 | 1,847 | $24 | 0.8% |
Your US homepage handles 72% of all ad traffic. Homepage performance is the single biggest lever for improving your US post-click quality.
EU (68 unique URLs, last 7 days)
| Landing Page | Clicks | Impressions | Cost | CTR |
|---|---|---|---|---|
| eu.juicymarbles.com (HTTP) | 426 | 11,788 | $36 | 3.6% |
| eu.juicymarbles.com (HTTPS) | 87 | 5,368 | $60 | 1.6% |
| /collections/all-products | 48 | 18,140 | $37 | 0.3% |
| Recipe blog pages (dozens) | 0-2 each | varies | minimal | <1% |
Three Problems Found
HTTP vs HTTPS Split (EU)
Your EU homepage is receiving ad traffic on both http://eu.juicymarbles.com (426 clicks) and https://eu.juicymarbles.com (87 clicks). Google treats these as two different landing pages, splitting your quality signal. The HTTP version forces a redirect to HTTPS, adding 100-300ms of load time.
https://eu.juicymarbles.com. This alone could improve your EU Post-Click Quality score.URL Fragmentation (EU)
EU has 68 unique landing page URLs but the vast majority receive 0-2 clicks each. Many are recipe blog pages that attract impressions but almost zero clicks. This fragmentation spreads your quality signal thin and makes it harder for Google to build a quality assessment.
Post-Click Quality Below Average (EU)
Your EU Quality Score data shows Post-Click Quality averaging 2.9/4. The keywords dragging this down are your generic terms where the landing page doesn't immediately address the search intent.
Page Speed Data (Live PageSpeed Insights - March 2026)
Mobile Performance
| Page | Performance | LCP | FCP | TBT | CLS |
|---|---|---|---|---|---|
| US Homepage | 41/100 | 19.7s | 6.8s | 586ms | 0.04 |
| EU Homepage | 55/100 | 20.6s | 3.2s | 264ms | 0.00 |
| US All Products | 47/100 | 34.2s | 6.4s | 374ms | 0.00 |
| EU All Products | 57/100 | 22.1s | 2.7s | 311ms | 0.02 |
Desktop Performance
| Page | Performance | LCP | FCP | TBT | CLS |
|---|---|---|---|---|---|
| US Homepage | 67/100 | 1.1s | 0.7s | 375ms | 0.20 |
| EU Homepage | 82/100 | 2.7s | 0.7s | 91ms | 0.01 |
| US All Products | 65/100 | 3.1s | 0.7s | 307ms | 0.00 |
| EU All Products | 72/100 | 2.4s | 0.7s | 252ms | 0.07 |
The headline number: your mobile LCP is 19-34 seconds
Google considers anything over 4 seconds "Poor." Your pages take 5-8x longer than Google's "Poor" threshold. This is the single biggest drag on your Post-Click Quality score. Every additional second of load time reduces conversion rate ~7%. A 19-second LCP means roughly half your paid mobile visitors have already bounced before seeing your products.
Other Scores (both stores perform well)
| Category | US Homepage | EU Homepage |
|---|---|---|
| Accessibility | 93/100 | 93/100 |
| SEO | 100/100 | 100/100 |
| Best Practices | 77/100 | 96/100 |
Cost Impact of Landing Page Quality
| If Post-Click Quality Improved | QS Impact | CPC Change | Weekly Savings |
|---|---|---|---|
| 2.9 to 3.5 (solidly "Above Average") | +0.5 to +1 QS point | ~8-13% reduction | ~$13-21 |
| 2.9 to 4.0 ("Well Above Average") | +1 to +1.5 QS points | ~13-22% reduction | ~$21-36 |
Appendix: Data Sources
| Source | File | Period |
|---|---|---|
| Meta Ads | reports/meta/all_ads_oct2024_mar2026.csv | Oct 2024 - Mar 2026 |
| Google Ads US QS | reports/google-ads/us/quality_score_last_7_days.csv | Last 7 days |
| Google Ads EU QS | reports/google-ads/eu/quality_score_last_7_days.csv | Last 7 days |
| Google Ads US Keywords | reports/google-ads/us/keyword_performance_last_7_days.csv | Last 7 days |
| Google Ads EU Keywords | reports/google-ads/eu/keyword_performance_last_7_days.csv | Last 7 days |
| Google Ads US Landing Pages | reports/google-ads/us/landing_page_last_7_days.csv | Last 7 days |
| Google Ads EU Landing Pages | reports/google-ads/eu/landing_page_last_7_days.csv | Last 7 days |
Note: Google Ads data covers a 7-day window. Meta data covers 18 months (Oct 2024 - Mar 2026). Conversion attribution may differ between platforms. ROAS figures use platform-reported purchase values. FC identification: Ads containing "_FC_" in the ad name are attributed to Flat Circle agency.