Forensic Audit Report

February ROAS
Was Inflated

A data-driven investigation into the February 2026 conversion anomaly in the Vegamour Google Ads account.

02 — THE ANOMALY

Something happened
on February 9th

Conversions jumped 786% overnight while ad spend stayed flat. That's physically impossible in real advertising.

Normal Daily Conversions
~208
Feb 1–8 average
Peak Anomaly Conversions
2,665
Feb 12 — 12.8× normal
Spend Change
+8%
$6,712 → $7,358/day
Conv. Change
+786%
208 → 1,757/day

⚠ The Physical Impossibility

On Feb 12, the account recorded 2,665 conversions on approximately 2,500 clicks. That implies a 106% conversion rate — meaning every single click converted, and some converted twice. This is physically impossible.

03 — VISUAL PROOF

Conversion volume by day

Green bars = normal. Red bars = anomaly. The spike is unmistakable.

Feb 1
206
Feb 2
238
Feb 3
218
Feb 4
183
Feb 5
217
Feb 6
159
Feb 7
202
Feb 8
240
Feb 9
877
Feb 10
1,790
Feb 11
2,522
Feb 12
2,665
Feb 13
931
Feb 14
127
04 — THE COMPARISON

Spend barely changed.
Conversions exploded.

Baseline (Feb 1–8)
Avg Daily Spend
$6,712
Avg Daily Conversions
208
Avg Daily Revenue
$19,555
ROAS
291%
Anomaly (Feb 9–13)
Avg Daily Spend
$7,358 (+8%)
Avg Daily Conversions
1,757 (+786%)
Avg Daily Revenue
$50,682 (+159%)
ROAS
689% (inflated)

You cannot get 786% more conversions from 8% more spend. The extra conversions did not come from ads — they were imported into the tracking system from an external source (Shopify webhook batch import of historical transactions).

05 — CORROBORATION

Only vega Sales spiked.
Subscriptions didn't.

If this were a real sale event, all purchase types would spike together. Only the Shopify webhook action (vega Sales) was affected.

Datevega Sales Convvega Sales ValueSubscription ConvSub ValueFlag
Feb 8234$21,6905$158Normal
Feb 9871$34,7746$186⚠️ 3.7× spike
Feb 101,790$50,0020$0⚠️ 7.7× spike
Feb 112,521$61,7231$54⚠️ 10.8× spike
Feb 122,664$63,6321$164⚠️ 11.4× spike
Feb 13918$42,72413$152⚠️ 3.9× spike
Feb 14126$14,2091$125Normal

Key finding: The Subscription action tracked normally throughout the anomaly (0–13 conv/day, consistent with baseline). Only the vega Sales action — which receives data via Shopify webhook — spiked. This confirms the source was a Shopify data import, not real ad-driven purchases.

06 — FULL FEBRUARY DATA

Every day. Every number.

DateSpendConvConv ValueROASConv/$1K
Feb 1$8,567206$22,552263%24.0
Feb 2$7,291238$20,528282%32.7
Feb 3$7,753218$23,086298%28.2
Feb 4$5,654183$17,938317%32.3
Feb 5$5,768217$17,347301%37.6
Feb 6$6,359159$15,690247%25.0
Feb 7$5,854202$17,390297%34.4
Feb 8$6,451240$21,911340%37.2
Feb 9$5,962877$34,960586%147.1
Feb 10$6,7231,790$50,002744%266.3
Feb 11$8,1072,522$61,777762%311.2
Feb 12$8,9792,665$63,795711%296.8
Feb 13$7,017931$42,876611%132.6
Feb 14$7,856127$14,334182%16.2
Feb 15$3,678236$27,609751%64.1
Feb 16$5,437210$23,003423%38.6
Feb 17$7,568175$18,828249%23.1
Feb 18$8,614184$22,156257%21.3
Feb 19$6,880159$17,745258%23.1
Feb 20$8,804146$18,508210%16.5
Feb 21$8,266231$24,264294%28.0
Feb 22$9,086189$21,459236%20.7
Feb 23$5,669198$18,763331%34.9
Feb 24$8,138214$26,483325%26.3
Feb 25$10,007228$26,750267%22.8
Feb 26$8,695174$19,930229%20.0
Feb 27$9,698215$24,438252%22.2
Feb 28$9,176148$16,116176%16.1
07 — GOOGLE ADS DASHBOARD

The spike visible in the
Google Ads dashboard

Cost (blue) stays flat. Conversions (orange) and Revenue (green) spike dramatically — then crash back to normal.

Google Ads dashboard showing conversion spike
Google Ads Overview — the spike around mid-February is clearly visible. Cost (blue line) remains flat while conversions and revenue spike then return to baseline.
08 — THE VERDICT

Reported vs. Corrected

Reported February ROAS
351%

Includes ~$156K phantom revenue
from batch-imported historical transactions

Corrected February ROAS
273%

Based on actual ad-driven purchases
excluding the Feb 9–13 batch import

Summary of Findings
  • 1
    Spend increased only 8% during the anomaly window — conversions increased 786%
  • 2
    Implied conversion rate of 106%+ is physically impossible
  • 3
    Only the vega Sales action (Shopify webhook) spiked — Subscription tracking was unaffected
  • 4
    The 5-day ramp-up → peak → ramp-down pattern is consistent with batch data processing
  • 5
    Average order value dropped from $96 to $32 during the anomaly — consistent with historical subscription renewals being backfilled
  • 6
    Total phantom revenue attributed: approximately $155,634