Melleka was brought in to manage Vegamour's Google Ads account. Within the first week, we identified critical data integrity issues that were masking true performance, fixed them, and established a clean measurement framework focused on real purchase revenue only.
The account had 9+ conversion actions feeding into reporting, many of which were not actual purchases. This inflated the reported ROAS by approximately 5.2x, making the account appear far more profitable than it actually was.
| Conversion Action | Status | Category | Revenue | Conversions |
|---|---|---|---|---|
| vega Sales | PRIMARY | Purchase (Shopify) | $79,083 | 701 |
| Purchase — Subscription | PRIMARY | Subscription | $29,273 | 299 |
| Add to Cart GTM | INFLATOR | Add to Cart | $208,494 | 2,638 |
| Begin Checkout GTM | INFLATOR | Begin Checkout | $74,343 | 740 |
| Purchase GTM | DUPLICATE | Purchase (duplicate) | $73,809 | 653 |
| vegaLASH GA4 Purchase | DUPLICATE | Purchase (duplicate) | $53,856 | 482 |
| Purchase — Straight Sale No Subs | DUPLICATE | Purchase (duplicate) | $47,239 | 433 |
| REAL Revenue (vega Sales + Subscription only) | $108,356 | 1,000 | ||
| INFLATED Total (all actions counted) | $566,097 | 5,946 | ||
For every $1 of real revenue, the old reporting showed $5.22. An account appearing to generate $566K was actually generating $108K. Bidding algorithms were optimizing toward add-to-cart events instead of actual purchases, wasting budget on users who browse but never buy.
Using real revenue only (vega Sales + Subscriptions), here is every day since we began managing the account.
| Date | Day | Spend | Revenue | ROAS | Grade |
|---|---|---|---|---|---|
| Mar 17 | Tue | $6,590 | $16,699 | 2.53x | B+ |
| Mar 18 | Wed | $4,314 | $14,238 | 3.30x | A |
| Mar 19 | Thu | $3,613 | $18,968 | 5.25x | A |
| Mar 20 | Fri | $4,144 | $14,533 | 3.51x | A |
| Mar 21 | Sat | $4,924 | $15,668 | 3.18x | A |
| Mar 22 | Sun | $4,895 | $16,936 | 3.46x | A |
| Mar 23 | Mon | $5,768 | $14,984 | 2.60x | B+ |
| Mar 24 | Tue | $5,222 | $20,387 | 3.90x | A |
| Mar 25 | Wed | $3,798 | $15,363 | 4.04x | A |
| Mar 26 | Thu | $3,179 | $24,540 | 7.72x | A |
| Mar 27 | Fri | $5,131 | $17,116 | 3.34x | A |
| Mar 28 | Sat | $5,113 | $11,838 | 2.32x | B |
| Mar 29 | Sun | $4,990 | $15,433 | 3.09x | A |
| Mar 30 | Mon | $5,014 | $9,799 | 1.95x | C |
| Mar 31 | Tue | $5,377 | $14,497 | 2.70x | B+ |
| March Total | $72,072 | $240,999 | 3.34x | A | |
| Apr 1 | Wed | $5,217 | $11,990 | 2.30x | B |
| Apr 2 | Thu | $4,934 | $11,870 | 2.41x | B |
| Apr 3 | Fri | $5,086 | $24,040 | 4.73x | A |
| Apr 4 | Sat | $5,736 | $12,400 | 2.16x | B |
| Apr 5 | Sun | $7,083 | $12,990 | 1.83x | C |
| Apr 6 | Mon | $6,428 | $12,076 | 1.88x | C |
| Apr 7 | Tue | $5,815 | $11,645 | 2.00x | B |
| Apr 8 | Wed | $5,666 | $10,476 | 1.85x | C |
| Apr 9 | Thu | $7,103 | $13,430 | 1.89x | C |
| Apr 10 | Fri | $7,540 | $14,543 | 1.93x | C |
| Apr 11 | Sat | $8,416 | $14,025 | 1.67x | C |
| Apr 12 | Sun | $6,117 | $13,990 | 2.29x | B |
| Apr 13 | Mon | $6,373 | $11,445 | 1.80x | C |
| Apr 14 | Tue | $5,745 | $11,804 | 2.05x | B |
| Apr 15 | Wed | $7,073 | $11,741 | 1.66x | C |
| Apr 16 | Thu | $6,978 | $11,572 | 1.66x | C |
| April Total | $101,310 | $210,040 | 2.07x | B | |
Performance by campaign from Apr 11–17, ranked by spend. Real revenue only.
| Campaign | Status | Spend | Revenue | ROAS | Verdict |
|---|---|---|---|---|---|
| Branded Search USA | ENABLED | $16,858 | $47,911 | 2.84x | Making Money |
| PMax Hair Growth Serums | ENABLED | $7,918 | $13,045 | 1.65x | Marginal |
| Branded Shopping USA | ENABLED | $6,226 | $9,993 | 1.61x | Marginal |
| Nonbranded Shopping USA | ENABLED | $4,082 | $2,103 | 0.52x | Burning Money |
| PMax Shampoo Kits | ENABLED | $2,794 | $6,495 | 2.32x | Making Money |
| NB Combo Search Exact | ENABLED | $2,082 | $967 | 0.46x | Burning Money |
| NB Lash Search | ENABLED | $1,647 | $1,228 | 0.75x | Burning Money |
| NB Broad AI-Max | PAUSED | $894 | $177 | 0.20x | Paused |
| NB Brow Search | ENABLED | $856 | $134 | 0.16x | Burning Money |
| Remarketing (Demand Gen) | ENABLED | $744 | $219 | 0.29x | Burning Money |
Within the first days of managing this account, we identified and fixed several issues that were actively wasting budget and corrupting data.
Identified that add-to-cart ($208K), begin-checkout ($74K), and 3 duplicate purchase tags ($175K) were inflating reported ROAS by 5.2x. Established vega Sales (Shopify webhook) as the single truth metric. All other actions moved to secondary/observation only.
March at $4,805/day delivered 3.34x ROAS. When spend jumped to $6,332/day in April, ROAS dropped to 2.07x. Revenue stayed flat at ~$13K/day regardless of how much was spent. This revealed clear diminishing returns — the market has a ceiling for daily purchase volume that additional spend cannot break through.
Non-branded campaigns (NB Brow at 0.16x, NB Broad at 0.20x, Remarketing at 0.29x, NB Combo Exact at 0.46x, NB Shopping at 0.52x) are collectively burning ~$9,400/week with negative returns. Pausing or heavily cutting these recovers budget that can be redirected to profitable branded and PMax campaigns.
Built and deployed an automated daily ROAS brief sent to the team every morning. Reports real revenue only, campaign-by-campaign breakdown, and specific recommendations. No more guessing about performance — the team has a single source of truth every day.
The math is straightforward. Revenue is relatively stable at $12-14K/day. The lever we control is spend. Reducing waste spend while maintaining revenue-generating campaigns gets us back to 3.0x.
| Action | Campaign | Weekly Savings | Impact |
|---|---|---|---|
| Pause | NB Brow Search (0.16x) | ~$856 | Zero revenue loss |
| Pause | Remarketing / Demand Gen (0.29x) | ~$744 | Minimal revenue loss |
| Cut 50% | Nonbranded Shopping (0.52x) | ~$2,041 | Test at lower spend |
| Cut 50% | NB Combo Exact (0.46x) | ~$1,041 | Test at lower spend |
| Scale +20% | PMax Shampoo Kits (2.32x) | +$559 | High ROAS, room to grow |
| Net Weekly Budget Recovery | ~$4,123 | ~$589/day savings | |
Projections based on current trajectory, planned optimizations, and the account's demonstrated revenue ceiling. All numbers use real revenue only.