Saved: 2026-03-25T23:56:26.975646+00:00
Model: gpt-5.4
Estimated input/output tokens: 27,293 / 11,285
CLIENT ASK Provide specific Google Ads optimizations for SipJeng based only on the 3 attached reports, with the stated goal of achieving the lowest CPA for purchase conversions. Analysis type is conversion; preferred output style is operator. PROVIDED EVIDENCE 1) Landing page report CSV - Title/date range: “Landing page report ‘September 25, 2025 - March 23, 2026’” - Fields include: Landing page, Selected by, Clicks, Impr., CTR, Avg. CPC, Cost, Conversions - Contains URL-level landing page performance plus account/channel totals 2) Channel performance CSV - Title/date range shown as: “Search terms insight report ‘September 25, 2025 - March 23, 2026’” - Despite title, content is channel/campaign performance by channel - Fields include: Channels, Status, Campaigns, Impr., Clicks, Interactions, Conversions, Conv. value, Cost, Results, Results value - Includes Google Search, Search partners, GDN, YouTube, Gmail, Discover, Maps - Includes multiple paused historical PMax/catch-all campaigns and one active “Cube | New Pmax” 3) Search terms report CSV - Title/date range: “Search terms report ‘September 25, 2025 - March 23, 2026’” - Fields include: Search term, Match type, Added/Excluded, Campaign, Ad group, Avg. CPM, Clicks, Impr., CTR, Avg. CPC, Cost, Campaign type, Conv. rate, Conversions, Cost / conv. - Search term file is truncated in the provided text, so not all rows are visible EXTRACTED FACTS Account / totals - Landing page totals: - Total landing pages: 3,120 clicks, 147,440 impressions, 2.12% CTR, $2.88 avg CPC, $8,984.10 cost, 351.49 conversions - Total account: 3,343 clicks, 147,440 impressions, 2.27% CTR, $2.97 avg CPC, $9,928.11 cost, 351.49 conversions - Total Performance Max: 499 clicks, 30,413 impressions, 1.64% CTR, $0.79 avg CPC, $391.91 cost, 1.00 conversion - Total Search: 2,844 clicks, 117,027 impressions, 2.43% CTR, $3.35 avg CPC, $9,536.20 cost, 350.49 conversions - Approx blended CPA from account total: $9,928.11 / 351.49 = about $28.25 - Approx search CPA: $9,536.20 / 350.49 = about $27.21 - Approx PMax CPA from landing page totals: $391.91 / 1.00 = $391.91 Channel/campaign performance - Channel totals from source 2: - All campaigns total: 556,348 impressions, 3,973 clicks, 69,895 interactions, 126.33 conversions, $10,027.42 conv. value, $8,347.53 cost - Google Search total: 214,867 impressions, 1,877 clicks, 126.33 conversions, $10,027.42 conv. value, $7,309.65 cost - Search partners total: 222 impressions, 5 clicks, 0 conversions, $3.31 cost - Google Display Network total: 183,361 impressions, 1,702 clicks, 0 conversions, $492.40 cost - YouTube total: 157,826 impressions, 389 clicks, 0 conversions, $540.58 cost - Gmail total: 72 impressions, 0 clicks, 0 conversions, $1.58 cost - Discover / Maps: 0 - There is a mismatch between source 1 and source 2 conversion totals: - Source 1 total conversions = 351.49 - Source 2 total conversions = 126.33 - Likely due to different conversion definitions / attribution / included campaigns, but this is not explicitly documented Campaign-level notable rows from source 2 - Paused “Cube_Catch All_OCT” on Google Search: - 135,613 impressions, 1,418 clicks, 94.88 conversions, $9,153.13 conv. value, $5,334.65 cost - Rough CPA ≈ $56.23 - Paused “Cube_30Dec_CatchAll_Pmax” on Google Search: - 72,373 impressions, 300 clicks, 28.44 conversions, $715.66 conv. value, $1,251.03 cost - Rough CPA ≈ $43.99 - Purchase results shown separately: Purchase 7.01 with purchase value $371.90 - Active “Cube | New Pmax” on Google Search: - 1,618 impressions, 63 clicks, 1.00 conversion, $23.09 conv. value, $198.46 cost - Rough CPA = $198.46 - Purchase: 1.00 - Active “Cube | New Pmax” on GDN: - 24,629 impressions, 429 clicks, 0 conversions, $154.22 cost - Active “Cube | New Pmax” on YouTube: - 4,107 impressions, 5 clicks, 0 conversions, $36.98 cost - Paused “Cube_Pmax” on Google Search: - 2,661 impressions, 81 clicks, 1.00 conversion, $26.00 conv. value, $481.72 cost - Rough CPA = $481.72 - Paused “Cube | PMax - Website Traffic” on Google Search: - 1,554 impressions, 11 clicks, 1.01 conversions, $109.55 conv. value, $30.16 cost - CPA ≈ $29.86 - But this appears to be mixed result types (add to cart / begin checkout / page view / purchase), not clean purchase-only - Strong evidence that Google Search drives all tracked conversions in source 2; display/video channels spend with zero conversions Landing pages with strongest purchase efficiency signals from source 1 - https://sipjeng.com/collections/best-sellers (ADVERTISER) - 791 clicks, 55,088 impressions, 1.44% CTR, $1.20 avg CPC, $951.15 cost, 207.65 conversions - Approx CPA ≈ $4.58 - https://shop.sipjeng.com/ (ADVERTISER) - 438 clicks, 17,308 impressions, 2.53% CTR, $3.30 avg CPC, $1,444.84 cost, 38.50 conversions - Approx CPA ≈ $37.53 - https://shop.sipjeng.com/shop/ (ADVERTISER) - 872 clicks, 68,994 impressions, 1.26% CTR, $3.71 avg CPC, $3,231.88 cost, 29.33 conversions - Approx CPA ≈ $110.16 - https://try.sipjeng.com/ (ADVERTISER) - 728 clicks, 21,337 impressions, 3.41% CTR, $3.85 avg CPC, $2,802.50 cost, 44.00 conversions - Approx CPA ≈ $63.69 - https://sipjeng.com/products/thc-infused-jeng-and-tonic (AUTOMATIC) - 23 clicks, 450 impressions, 5.11% CTR, $5.05 avg CPC, $116.05 cost, 6.00 conversions - Approx CPA ≈ $19.34 - https://sipjeng.com/products/non-alcoholic-thc-drinks? no exact URL: - Actual URL: https://sipjeng.com/collections/non-alcoholic-thc-drinks - 18 clicks, 507 impressions, 3.55% CTR, $3.26 avg CPC, $58.71 cost, 4.00 conversions - Approx CPA ≈ $14.68 - https://sipjeng.com/about page (AUTOMATIC) - 6 clicks, 19 impressions, 31.58% CTR, $5.53 avg CPC, $33.15 cost, 2.00 conversions - Approx CPA ≈ $16.58 - Very small sample - https://sipjeng.com/ homepage (AUTOMATIC) - 30 clicks, 194 impressions, 15.46% CTR, $1.68 avg CPC, $50.45 cost, 2.00 conversions - Approx CPA ≈ $25.23 - https://shop.sipjeng.com/product/collection-sampler-6-pack/ (ADVERTISER) - 20 clicks, 13,454 impressions, 0.15% CTR, $4.98 avg CPC, $99.65 cost, 4.00 conversions - Approx CPA ≈ $24.91 - https://sipjeng.com/blogs/blog/alcohol-alternative-drinks-2025 (AUTOMATIC) - 225 clicks, 2,104 impressions, 10.69% CTR, $1.88 avg CPC, $423.97 cost, 10.00 conversions - Approx CPA ≈ $42.40 Landing pages with spend but zero conversions - https://sipjeng.com/products/thc-infused-paloma (AUTOMATIC): 8 clicks, $61.39, 0 conv - https://sipjeng.com/collections/cbd-infused-drinks (AUTOMATIC): 20 clicks, $77.91, 0 conv - https://shop.sipjeng.com/product/sweet-spot-pack/ (ADVERTISER): 2 clicks, $7.95, 0 conv - https://shop.sipjeng.com/about/ (ADVERTISER): 3 clicks, $24.38, 0 conv - https://shop.sipjeng.com/contact/ (ADVERTISER): 5 clicks, $20.05, 0 conv - https://shop.sipjeng.com/about/ (AUTOMATIC): 1 click, $2.95, 0 conv - https://shop.sipjeng.com/contact/ (AUTOMATIC): 0 clicks, 2,520 impressions, 0 conv - https://shop.sipjeng.com/product/summer-starter-pack/ (ADVERTISER): 1 click, $16.61, 0 conv - https://sipjeng.com/collections/functional-beverages (AUTOMATIC): 6 clicks, $35.39, 0 conv - Multiple blog/article pages have spend and 0 conversions Search terms / intent observations from source 3 - Branded term “sipjeng” in campaign Cube_Search_W, ad group “Ad group 1”: - 2 clicks, 2 impressions, 100% CTR, $0.17 avg CPC, $0.34 cost, 14.00 conversions, cost/conv $0.02 - This is an extreme outlier and likely reflects data/attribution inflation or micro-conversion mapping; should not be trusted at face value for purchase CPA decisions without clarification - “mocktails”: - 1 click, 36 impressions, $0.85 cost, 1.00 conversion, cost/conv $0.85 - Too small a sample - Non-brand waste examples with clicks and no conversions: - “hemp infused seltzer” 1 click, $3.46, 0 conv - “tost discount code” 1 click, $7.43, 0 conv - “cbd drinks 50 mg” 1 click, $10.35, 0 conv - “nootropic drinks to replace alcohol” 4 clicks, 8 impressions, $9.03, 0 conv - “relaxing drinks instead of alcohol” 1 click, $3.75, 0 conv - Many visible search terms are competitor terms / generic exploratory / recipe / informational: - competitor/brand-like: shimmerwood beverages, gaba spirits, melati drinks, wunder drink, cycling frog drinks, little saints negroni, seth rogen seltzer, where to buy de soi, ohho drinks, nowadays drink near me, cann tonics, cann social tonic packets, athletic brewing seltzer, breez/brez-like queries, etc. - informational/recipe-like: valentines cocktail recipes, greyhound drink, freezer old fashioned, skeleton key drink, strawberry hennessy, mocktails with club soda, drink recipes non alcoholic, making a mocktail, monte carlo cocktail, hunch punch, campechana drink, etc. - Visible evidence suggests broad/AI Max traffic is matching to many irrelevant or low-intent queries OBSERVED METRICS Primary measurable figures - Account total CPA (source 1): ~$28.25 - Search CPA (source 1): ~$27.21 - PMax CPA (source 1): ~$391.91 - Channel total CPA (source 2 using total conversions): ~$66.08 - Google Search CPA (source 2): ~$57.86 - GDN / YouTube CPA in source 2: no conversions Best visible landing page CPAs with meaningful volume - /collections/best-sellers (ADVERTISER): ~$4.58 CPA on 207.65 conv - /try.sipjeng.com/: ~$63.69 CPA on 44 conv - /shop.sipjeng.com/ (ADVERTISER): ~$37.53 CPA on 38.5 conv - /shop.sipjeng.com/shop/ (ADVERTISER): ~$110.16 CPA on 29.33 conv - /blogs/blog/alcohol-alternative-drinks-2025: ~$42.40 CPA on 10 conv - /products/thc-infused-jeng-and-tonic: ~$19.34 CPA on 6 conv - /collections/non-alcoholic-thc-drinks: ~$14.68 CPA on 4 conv - /shop.sipjeng.com/product/collection-sampler-6-pack/ (ADVERTISER): ~$24.91 CPA on 4 conv Notable CTRs - /blogs/blog/alcohol-alternative-drinks-2025: 10.69% CTR - /sipjeng.com/ homepage (AUTOMATIC): 15.46% CTR - /pages/about (AUTOMATIC): 31.58% CTR but tiny volume - /collections/best-sellers (ADVERTISER): 1.44% CTR on large volume - /shop.sipjeng.com/shop/ (ADVERTISER): 1.26% CTR on large volume - /try.sipjeng.com/ (ADVERTISER): 3.41% CTR Contradictions / data quality flags - Conversion totals differ sharply across reports: - 351.49 in landing page report vs 126.33 in channel report - Source 2 “Results” mixes actions like Page View, Add to cart, Begin checkout, Purchase - Source 1 “Conversions” likely not guaranteed to be purchase-only despite client goal - Search term report includes implausible conversion rates: - “sipjeng” 14 conversions from 2 clicks, 700% conv rate - indicates conversion aggregation / fractional attribution / mixed conversion actions GAPS/UNCERTAINTY - No screenshots were provided; only CSV text exports - The “3 reports” are not enough to isolate purchase-only CPA confidently because: 1) conversions appear mixed across actions in source 2 2) source 1 does not explicitly say conversions = purchases 3) source 3 shows fractional and >100% conversion rates, suggesting non-last-click or mixed-action attribution - Search term file is truncated, so full waste/opportunity review is incomplete - No campaign budget data, bid strategy settings, asset group data, geo/device/daypart/audience performance, or actual purchase-only conversion action settings - No campaign names in landing page report, so URL performance cannot be tied to exact campaign/ad group structure - No explicit split of brand vs non-brand performance except partial visible search terms - No ROAS target / margin threshold / allowable CPA target provided - No evidence on conversion tracking hygiene; likely a major issue given inconsistent totals and multi-action reporting - No evidence whether “purchase conversion” is primary account goal or whether page view / checkout are included as primary conversions RECOMMENDED ANALYSIS ANGLE 1) Lead with a tracking caveat: - Before making aggressive scaling decisions, flag that the data is inconsistent for purchase-only CPA optimization. Recommend verifying that only Purchase is a primary optimization/conversion goal in Google Ads, and demote Page View / Add to cart / Begin checkout if they are contaminating optimization/reporting. 2) Strongest optimization thesis from evidence: - Shift traffic concentration toward proven purchase-efficient landing pages: - prioritize /collections/best-sellers - test/promote /collections/non-alcoholic-thc-drinks - test/promote /products/thc-infused-jeng-and-tonic - be cautious but consider /shop.sipjeng.com/product/collection-sampler-6-pack/ - Reduce or remove spend to weak destinations: - /shop.sipjeng.com/shop/ - product/about/contact pages with spend and no conversions - low-performing blog and informational pages unless they are intentionally top-funnel and excluded from purchase campaigns 3) Channel/campaign control recommendation: - Search is the only channel with conversions in source 2; GDN/YouTube spend is non-converting. - Recommend constraining or excluding non-search inventory within active PMax if possible, or reducing/pausing PMax until purchase-only signal quality is fixed. - Search partners show negligible volume and zero conversions; likely safe to disable. 4) Query hygiene recommendation: - Tighten match types and add negatives for competitor, recipe, informational, discount-code, near-me, and generic beverage terms shown in source 3. - Build negatives from visible waste terms such as “discount code,” recipe/cocktail names, competitor brands, and broad beverage discovery queries. 5) Structure recommendation: - Separate brand vs non-brand clearly. - Route high-intent commercial non-brand traffic to best-performing collection/product pages, especially /collections/best-sellers. - If keeping content/blog terms, isolate them in a separate lower-budget campaign with stricter CPA guardrails. 6) Decision criteria likely to matter to client: - Lowest purchase CPA first - Secondary: retaining volume if CPA remains acceptable - Preference for concrete account actions: pause, reduce, exclude, redirect traffic, add negatives, split campaigns, change landing pages, verify conversion goal settings
Your stated goal is lowest CPA for purchase conversions. Based on the three exports, the clearest operating conclusion is this:
Sanity check on the arithmetic:
Those totals do not agree. Source 1 shows 351.49 conversions; source 2 shows 126.33. Source 2 also mixes Page View, Add to cart, Begin checkout, and Purchase in “Results.” Source 3 contains impossible-looking lines like 14 conversions from 2 clicks on “sipjeng.” So the directional read is strong, but the purchase-only CPA confidence is weak until conversion settings are cleaned up.
The easiest waste cuts from the evidence are outside pure Search and inside loose query matching.
If the objective is lowest purchase CPA, there is no support here for continuing to feed display/video inventory.
Against those, https://sipjeng.com/collections/best-sellers shows $951.15 / 207.65 ≈ $4.58 CPA, which is dramatically lower. Even allowing for attribution noise, that gap is too large to ignore.
The visible query set shows broad or AI Max matching into low-intent traffic. Specific waste examples with spend and zero conversions:
There are also many visible competitor and informational/recipe queries: shimmerwood beverages, gaba spirits, melati drinks, wunder drink, cycling frog drinks, little saints negroni, where to buy de soi, greyhound drink, freezer old fashioned, valentines cocktail recipes, and similar terms. That is not lowest-CPA traffic.
If you want lowest CPA, scale should mean reallocating traffic to the destinations and intents that already show the best purchase-efficiency signals in the exports.
This is the strongest signal in the entire dataset with meaningful volume. I would make this the default landing page for high-intent non-brand search unless a tighter product/category page clearly beats it for the exact query theme.
These are promising routing tests, but they are not yet as proven as /collections/best-sellers because the click volume is much lower.
Right now, a lot of spend is going to weaker destinations:
For purchase CPA minimization, the immediate bet is to redirect existing paid search clicks away from those weaker pages and toward /collections/best-sellers first, then test the category/product pages above by tighter query intent.
Operator action: if your only goal is lowest purchase CPA, pause Cube | New Pmax until purchase tracking is verified, or reduce it to minimum while you shift budget into standard Search where intent and query control are stronger.
Small sample, but there is no evidence of value. Turn it off for CPA control.
The reports support a cleaner Search structure:
The search-term export shows branded and non-branded traffic mixed with loose matching behavior. Lowest-CPA management requires separate control. Brand CPA will almost always behave differently from non-brand.
Historical paused campaigns did convert, but not at your best visible CPA:
Those are better than the active PMax read, but still much worse than the best landing-page route. If you relaunch anything, do it as tighter Search campaigns with cleaner negatives and controlled landing pages, not broad catch-all.
Add these as campaign-level negatives in non-brand purchase campaigns unless you intentionally want competitor or informational traffic:
Examples directly supported by visible terms:
Visible competitor-style terms in the export include:
If you want pure lowest CPA, do not let the non-brand campaign absorb this traffic. Add exact and phrase negatives for the visible competitor set.
The export shows Broad match and AI Max surfacing weak-intent traffic. For purchase efficiency:
Do not keep routing broad traffic into /shop/ when that page is showing about $110.16 CPA in the report.
That is not a reliable purchase signal. Keep brand separate, but do not use that line as evidence that brand is infinitely efficient.
This is the strongest action supported by the exports. Compared with the current main alternatives:
Even if conversion definitions are messy, the relative gap is substantial enough to act on.
Exclude or stop using these in purchase-focused campaigns where they have spend and no visible conversions:
One blog page did show conversions:
That is not terrible, but it is still much weaker than /collections/best-sellers. If you keep content traffic, isolate it in a separate campaign and do not let it dilute your purchase-focused search budget.
Specific tests supported by the data:
The evidence is directionally useful but not clean enough to claim true purchase-only CPA with confidence.
Before you trust any automated bidding or scale decisions for “lowest purchase CPA,” confirm that only Purchase is set as the primary conversion goal used for bidding in the purchase campaigns. If Page View, Add to cart, or Begin checkout are included as primary goals, that is likely contaminating optimization.
You asked for specific Google Ads optimizations based only on the three attached reports, with the goal of the lowest CPA for purchase conversions.
Based on those reports, the highest-confidence actions are:
VERDICT revise SCORES request_match: 4 evidence_grounding: 3 detail_sufficiency: 4 actionability: 4 clarity: 4 overall: 3 WHAT WORKED - Strong alignment to the client goal of lowest CPA and mostly stayed within the 3 provided reports. - Correctly surfaced the biggest high-confidence themes from the evidence: - Search is the only channel with recorded conversions in source 2. - GDN/YouTube/Gmail/Search Partners show zero conversions in source 2. - /collections/best-sellers is the strongest visible landing page by CPA and volume in source 1. - Active Cube | New Pmax looks poor on the provided data. - Tracking quality is a major caveat because source totals conflict and source 2 mixes result types. - Arithmetic shown for major CPA figures is correct: - account CPA ≈ $28.25 - Search CPA from source 1 ≈ $27.21 - PMax CPA ≈ $391.91 - Google Search CPA from source 2 ≈ $57.86 - The output is operator-style and gives concrete actions, which matches the preferred style. - Good skepticism on the “sipjeng” row and on using mixed conversion data for bidding decisions. FAILURES - It overstates confidence in some recommendations that are not purchase-only proven. The biggest issue: many recommendations lean on source 1 landing page “Conversions,” but Agent 2 also acknowledges those conversions may not be purchases. That means statements like “make /collections/best-sellers the primary paid-search destination” are directionally plausible but too definitive for a purchase-CPA brief without stronger qualification. - It treats PMax as if it can be cleanly assessed by channel in a way that implies direct control over inventory (“pause or sharply restrict non-search inventory inside PMax / cross-network delivery”). The evidence shows channel rows for the campaign, but the recommendation should be framed more carefully: you can pause/reduce PMax, not directly turn off GDN/YouTube inside standard PMax in the same way as campaign channel toggles. - Some recommendations assume campaign structure detail not provided in the reports: - “change default final URL from /shop.sipjeng.com/shop/ to /collections/best-sellers” - “ad groups sending to /try.sipjeng.com/” - “remove blog URLs ... from active final URL use” The landing page report does not tie URLs to campaigns/ad groups, so these are inferred implementation steps, not evidenced account facts. - It includes at least one unsupported search term example not present in Agent 1 handoff: “wims discount code first order.” That is a hallucination relative to the provided evidence. - It claims the search-term export shows “Broad match and AI Max” explicitly. Agent 1 only said visible evidence suggests broad/AI Max traffic and noted loose matching; the report fields listed do not include an “AI Max” field. This is too assertive. - It does not sufficiently prioritize the conversion-tracking issue as the gating item. For a purchase CPA mandate, verifying Purchase as the only primary bidding goal should arguably be the first and strongest recommendation before making large budget reallocations. - It could have been more skeptical about low-volume landing page wins: - /collections/non-alcoholic-thc-drinks: 18 clicks, 4 conversions - /products/thc-infused-jeng-and-tonic: 23 clicks, 6 conversions - /collection-sampler-6-pack/: 20 clicks, 4 conversions These are useful tests, but not strong proof for scale decisions. MISSED EVIDENCE - It did not explicitly mention the magnitude of the discrepancy between source 1 account total and source 2 all-campaign total costs/conversions beyond CPA mismatch. The broader mismatch matters: - source 1 total cost $9,928.11 vs source 2 cost $8,347.53 - source 1 conversions 351.49 vs source 2 conversions 126.33 This reinforces that cross-report decisioning is highly unstable. - It missed a stronger caution that source 2 “Results” includes mixed actions and purchase appears only on some campaign rows. That undermines comparing campaign “conversions” and “results” as if they are the same KPI. - It did not note that paused historical campaigns dominate the source 2 conversion data. That means using source 2 to recommend current budget shifts has a recency/availability problem. - It did not mention that Gmail had spend with zero clicks, which is another sign of reporting oddities / non-core channel noise. - It could have highlighted that the landing-page report includes account/channel totals and URL rows, but not campaign names, limiting any exact “redirect these ad groups” instruction. HALLUCINATION CHECK - Unsupported/invented evidence found: - “wims discount code first order” was not in the provided evidence. - Potential overreach: - “The export shows Broad match and AI Max” is stronger than the handoff supports. Match type exists in source 3, but AI Max was only inferred by Agent 1, not evidenced as a field/explicit label. - Several implementation details imply known routing from ad groups/campaigns to specific URLs, which the landing page report does not prove. - No major arithmetic hallucinations found in the CPA calculations shown. REVISION NEEDED - Remove unsupported term “wims discount code first order.” - Reframe all landing-page recommendations as tests or reallocations under uncertainty, not as purchase-proven truths. Example: “prioritize testing /collections/best-sellers as the default destination for high-intent non-brand traffic, given its much lower reported CPA in source 1, while recognizing conversions may not be purchase-only.” - Tighten the PMax recommendation wording: - valid: pause/reduce PMax because visible active PMax performance is weak and non-search delivery shows zero conversions - avoid implying direct exclusion of GDN/YouTube inside PMax unless specifically supported - Move conversion-goal hygiene to the top priority and make it the gating action before aggressive budget shifts. - Add clearer caveats on low-volume page tests and on the fact that landing page report lacks campaign/ad-group linkage. - Distinguish “source 1 conversion CPA” from “purchase CPA” more consistently throughout. FINAL QA SUMMARY This is a solid draft directionally, but not clean enough to approve as-is for a strict evidence-based QA standard. It mostly matches the ask and contains useful, concrete actions, but it overreaches in a few places, includes at least one hallucinated search term, and presents some non-purchase conversion findings too definitively for a purchase-CPA objective. The biggest correction needed is to make tracking validation the top gating recommendation and to soften any claims that treat source 1 landing-page conversions as proven purchases.
No human feedback saved yet.