United States · Liquor ecommerce · Google Ads, GA4 and Shopify
The customer’s last complete month before our documented access returned 3.80× in Google Ads. We cleaned conversion tracking, rebuilt campaign roles, and kept working when the ad dashboard improved before the store did. By June, the complete account reached 7.20× platform ROAS. From Q1 to Q2, Shopify total sales rose 10.7%.
We start with the purchase signal, campaign economics, and store reality—not a larger budget.
01 · The real problem
A better advertising number did not automatically mean a healthier ecommerce business. We had to make Google’s purchase signal trustworthy, then confirm that paid traffic and store sales moved in the same direction.
Legacy campaigns and several conversion actions made one blended number hard to trust.
Shopify reported $134,644.25 in Q1 total sales and displayed a 47% decline against the preceding equal-length holiday comparison.
We checked GA4 paid-source purchases and Shopify store outcomes before deciding what deserved more budget.
02 · What we changed
We did not apply one blanket optimization. We rebuilt the decision chain from conversion action to campaign role to store-level validation.
We stopped old purchase, checkout, and add-to-cart actions from competing with the purchase signal that should guide bidding.
Five actions removed within days of accessWe separated product discovery, brand demand, high-intent search, and Shopping so each budget could prove what it earned.
New Shopping campaign: 6.13× through May 27Two May tests used 38.0% of spend but produced only 13.5% of attributed value. We paused both instead of scaling the blended account.
Weak May test pool: 1.22× platform ROASThe structure gave us a useful failure signal. When a test returned too little, we could stop it without starving the campaign pool that had earned budget.
03 · What moved
We kept every system in its proper lane. Google Ads told us which campaigns earned spend. GA4 showed paid-source purchase behavior. Shopify showed what happened across the whole store.
04 · What it means
The account no longer forced one blended number to answer every question. We could scale the active engine, protect store-level gains, and address the smaller-basket constraint separately.
It produced almost all June attributed value and excluded the test that failed.
Platform revenue matters less when the store gives too much of it back.
The next tests should address bundles, product mix, offers, and onsite merchandising.
Questions a careful buyer should ask
The screenshot shows the three campaigns that remain enabled: $32,176.38 in attributed value from $4,254.72 in spend. We report 7.20× for the complete June account because that also includes one high-intent test we later paused after it returned 0.60×.
The last complete month before our documented access returned 3.80×. We first gained documented access on December 3, 2025, so we use November as the clean pre-access reference and keep December inside the entry period.
We do not make that claim. Google Ads reports platform attribution, GA4 reports paid-source website activity, and Shopify reports whole-store results across every channel. Their shared direction supports the story; it does not create a controlled incrementality experiment.
No result comes with a guarantee. The repeatable part is our decision system: clean the purchase signal, separate campaign roles, stop weak spend, and check platform results against analytics and store outcomes before scaling.
Google Ads platform ROAS equals platform-attributed conversion value divided by ad cost. We calculate the 7.20× headline across every campaign with June spend. The screenshot’s 7.56× covers only the three campaigns that remain enabled.
We recalculate Shopify’s Q1-to-Q2 changes from the exact values in both native dashboards. Shopify’s Q2 interface compares Apr 1–Jun 30 with Dec 31–Mar 31, so its rounded arrows differ slightly from our clean calendar-quarter comparison.
GA4 paid-source data uses the exact session source / medium google / cpc. Google Ads, GA4, and Shopify use different attribution rules, date boundaries, and revenue definitions. We do not force them to match or relabel whole-store sales as ad revenue.
