Google Ads optimization — baseline snapshot (2026-06-13)
Captured immediately BEFORE the conversion-goal cleanup + tROAS reset, so we can measure impact on real purchases / revenue / website over the following weeks.
Account state BEFORE
- Account-default biddable goals: PURCHASE, ADD_TO_CART, BEGIN_CHECKOUT, PAGE_VIEW, DEFAULT (all biddable=true)
- Sales-Performance Max-1 (22681118791): MaxConvValue, tROAS 9, budget $25/day
- Eid Adha 2026 - PMax (23867754667): MaxConvValue, no target, budget $5/day
- Fresh Phoenicia Bread - PMax (23867767621): MaxConvValue, no target, budget $5/day
- search-campaign / search-campaign #2: PAUSED
Last-30-day performance BEFORE (real = purchase-only)
| Campaign | Spend | Reported conv | Reported value | Reported ROAS | Real purchases | Real revenue | Real ROAS |
|---|---|---|---|---|---|---|---|
| Sales-Performance Max-1 | $422.98 | 61.4 | $2,467.47 | 5.83x | 6.94 | $347.19 | 0.82x |
| Fresh Phoenicia Bread | $171.34 | 12.0 | $894.26 | 5.22x | 3.97 | $239.03 | 1.40x |
| Eid Adha 2026 | $136.40 | 6.0 | $247.13 | 1.81x | 4.00 | $214.73 | 1.57x |
| TOTAL | $730.72 | 79.4 | $3,608.86 | 4.94x | 14.9 | $800.95 | 1.10x |
Changes applied 2026-06-13
- ADD_TO_CART + BEGIN_CHECKOUT customer conversion goals -> biddable=false (secondary). Purchase stays primary.
- tROAS set to 1.5 on all three enabled PMax campaigns (main was 9; other two had none).
Ramp plan to ROAS 5
1.5 (breakeven floor) -> 2.0 -> 2.5 -> 3.5 -> 5.0, ~2 weeks per rung, stepping up only as real ROAS holds.
How to re-measure
Run /tmp/ga-query.py with the same GAQL (real purchases via segments.conversion_action_name = "Google Shopping App Purchase"). Website side: google-analytics-pp-cli (sessions, conversions, revenue) before/after.
Website (GA4) baseline BEFORE — last 30 days (2026-05-14 .. 2026-06-13)
| Channel | Sessions | Transactions | Revenue | AOV |
|---|---|---|---|---|
| Cross-network (PMax/Google Ads) | 1,026 | 11 | $561.72 | $51.07 |
| Direct | 474 | 3 | $153.54 | $51.18 |
| Paid Social (Meta) | 215 | 0 | $0 | — |
| Organic Search | 126 | 2 | $80.49 | $40.25 |
| Organic Social | 66 | 0 | $0 | — |
| 53 | 2 | $127.49 | $63.75 | |
| Unassigned | 33 | 0 | $0 | — |
| Paid Search | 19 | 0 | $0 | — |
| Organic Shopping | 10 | 0 | $0 | — |
| Referral | 8 | 0 | $0 | — |
| TOTAL | ~2,030 | 18 | ~$923 | ~$51 |
Note: GA4 attributes 11 txn / $562 to Cross-network (PMax), vs Google Ads' own 14.9 real purchases / $801 for the same campaigns. The gap = cross-domain attribution loss (storefront www.babanuj.com -> checkout.babanuj.com). Google Ads conversion tracking (Shopify Google & YouTube app, server-side) is the more reliable purchase count.
What to watch after the change (re-measure ~2026-06-27, after the learning period)
HYPOTHESIS: with only Purchase biddable + tROAS 1.5, Google stops buying cheap add-to-cart/checkout-start clicks. Expect: Cross-network SESSIONS flat-or-down, but conversion RATE and revenue-per-session UP, and real (purchase-only) ROAS climbing from 1.10x toward 1.5x. Spend may dip short-term during relearning.