Ecommerce measurement and paid-growth case study
The customer came to us while Google Ads was spending without a dependable revenue signal. March captured effectively no value. April reported 0.17×. We repaired measurement, rebuilt the account, and reached 10.86× in May, 14.59× in June, and a 19.00× quarterly peak—then used the Q2 regression to find the next growth constraint.
Spend continued while value capture failed
May 8–9 marked the new reporting regime
Quarterly platform ROAS peaked in Q1
Q2 exposed the next allocation problem
The signal
March spent INR 227.8K while Google Ads captured effectively no conversion value. April spent INR 273.2K and reported only 0.17×, while GA4 attributed INR 3.54M to google / cpc and Shopify’s structured export recorded INR 6.83M in total sales. Those platforms should not match, but that mismatch and 22 zero-value Ads days proved the account could not give bidding a dependable commercial signal.
Effectively no captured value
22 of 30 days recorded zero value
Value signal stabilized on May 8–9
Double-digit efficiency held
Calendar months · value / cost
| Month | Spend | Platform ROAS | Evidence note |
|---|---|---|---|
| Mar ’25 | INR 227.8K | ≈0× | Effectively no captured value |
| Apr ’25 | INR 273.2K | 0.17× | Incomplete value signal |
| May ’25 | INR 307.3K | 10.86× | Measurement stabilized |
| Jun ’25 | INR 303.3K | 14.59× | Double-digit result sustained |
Apr 1–Jun 30, 2025
The rebuild
We did not stop after the tag started returning value. We used the new signal to split demand, rebuild campaigns, change bids and budgets, refresh assets, and scale the product and audience combinations that kept earning spend.
Apr 1–May 7 · INR 335.8K spend
6.12× reported on the changeover day
May 9–Jun 30 · INR 538.5K spend
Average daily spend rose 12.0% across the boundary while captured value per day rose 4,848.7%. That discontinuity confirms a measurement-regime change. It does not support a claim that actual business return increased 44.2× overnight.
Jul 1–Sep 30, 2025
The interruption
GA4 and Shopify both show a sharp August 17–19 break. GA4 recorded a 93.7% three-day session collapse and no revenue or transactions on August 18–19. Shopify’s quarter-level daily record also omits August 18–19. We did not label the dip a campaign failure. We isolated a cross-system interruption and kept the cause open: measurement, traffic, site, or checkout.
Q3 2025
Q3 2025
The scale
After Q3 stabilized at 15.10×, the account reached 16.78× in Q4 and 19.00× in Q1 2026. From Q2 2025 to the Q1 peak, spend rose 35.7% while platform-attributed value rose 191.9%.
Weighted conversion value / cost
| Quarter | Spend | Platform value | ROAS |
|---|---|---|---|
| Q2 2025 | INR 883.8K | INR 7.81M | 8.83× |
| Q3 2025 | INR 1.10M | INR 16.59M | 15.10× |
| Q4 2025 | INR 1.08M | INR 18.16M | 16.78× |
| Q1 2026 | INR 1.20M | INR 22.79M | 19.00× |
| Q2 2026 | INR 1.28M | INR 20.12M | 15.78× |
INR 1.20M spend
INR 1.28M spend
The next constraint
Q2 spend rose 6.3% while platform value fell 11.7%. New Search and Demand Gen tests spent INR 130.6K at 3.96×; the continuing core returned 17.13×. The core also softened from the Q1 peak, so pausing tests alone could not explain the whole regression. We widened the diagnosis to GA4 and Shopify.
INR 130.6K spend · INR 517.0K platform value
INR 1.145M spend · INR 19.61M platform value
19.00× in Q1 → 15.78× in Q2
Three dashboards, three jobs
Google Ads
Allocation signalGA4
Measurement signalShopify
Store signalFrom Q2 2025 to Q2 2026, GA4 recorded 68.0% more sessions, 42.9% more revenue, and 54.7% more distinct transaction IDs. Yet revenue per session fell 15.0%. From Q1 to Q2 2026, sessions rose 20.8% while GA4 revenue fell 9.7% and revenue per session fell 25.3%. The account had reach; it needed stronger monetization.
Q2 2026
| GA4 metric | Q2 2025 | Q2 2026 | Change |
|---|---|---|---|
| Sessions | 287,898 | 483,700 | +68.0% |
| GA4-reported revenue | INR 22.46M | INR 32.09M | +42.9% |
| Distinct transaction IDs | 4,047 | 6,261 | +54.7% |
| Returning-session share | 22.58% | 32.52% | +9.93 pp |
| Revenue per session | INR 78.02 | INR 66.34 | −15.0% |
Quarter-aligned exports show Q2 total sales 22.3% above the prior-year quarter and 46.8% more orders. Weighted sales per order fell 16.7%, so volume carried more of the growth. Against Q1 2026, Q2 total sales fell 17.5% while orders fell only 3.4% and weighted sales per order fell 14.5%. That moved the next brief from reach to monetization quality.
Baseline quarter
Latest quarter
| Store metric | Q2 2025 | Q2 2026 | Change |
|---|---|---|---|
| Total sales | INR 20.98M | INR 25.65M | +22.3% |
| Net sales | INR 20.66M | INR 25.05M | +21.2% |
| Orders | 4,435 | 6,510 | +46.8% |
| Weighted sales per order | INR 4,730 | INR 3,941 | −16.7% |
The export record counts 4,435 and 6,510 orders in the two quarters. The dashboard widgets display 4,651 and 7,277. We preserve that definition boundary instead of silently forcing the figures to match.
Questions buyers ask
We want the story to earn confidence before it earns a call. These answers define what the evidence proves and how we would approach the same problem again.
No. Both figures are Google Ads platform value divided by Google Ads cost, and the 0.17× period had incomplete value capture. We repaired measurement, then sustained and scaled the reported platform result. Shopify remains the separate store record.
Because bidding cannot optimize toward revenue it cannot see. We first separated purchase value from duplicate, zero-value, and upper-funnel actions; then we used the cleaner signal to rebuild campaigns and allocate budget.
The value signal became sustained across Brand, Performance Max, competitor, nonbrand Search, and Shopping. A new Search ad group began serving, and Shopping expanded. The evidence supports a measurement-regime break and account rebuild, not a one-day creative miracle.
They use different attribution rules, event definitions, and scopes. We use Google Ads to steer paid-media allocation, GA4 to diagnose measured traffic and revenue, and Shopify to read all-channel store outcomes.
Repeated purchase events inflated GA4’s Transactions metric. We use distinct IDs as the safer analytics count, although they still do not replace validated backend orders.
Because a high-performing account still needs diagnosis. New tests returned 3.96×, the core softened, GA4 monetization fell behind traffic, and Shopify order value weakened. Hiding that quarter would hide the work that matters next.
The evidence shows a shared interruption in GA4 and Shopify, but it does not isolate the cause. We keep measurement, traffic, site availability, and checkout failure open until direct operational evidence resolves them.
No. Market demand, economics, inventory, creative, offer, site experience, and measurement all affect performance. We can guarantee a disciplined diagnosis, clear reporting boundaries, and an evidence-led scaling process.
Google Ads + GA4 + Shopify
Bring us the dashboards that disagree. We will trace the measurement gaps, rebuild the acquisition system, and show you what must improve before the next budget increase.
Find My Next Constraint