From zero recorded sales to the cheapest channel

From zero recorded sales to the cheapest channel

Online Sales · Beauty brand

A direct-to-consumer brand makes body makeup, and most of its buyers are over 45. Meta was the engine. Google Ads was running too, and spending, and reporting nothing at all. We reviewed both accounts, fixed what was broken and turned the findings into a plan.

0 → 115purchases recorded by Google Adswithin weeks of the fixes
$13.87Google cost per purchaseless than half of Meta’s $29.95
≈3.9×Google return after the rebuild
1,276Meta purchases in the quarter reviewed$67.5K in sales

The situation

An account that spent and never reported

The Google audit read like a checklist of small leaks: ads pointing to product pages that no longer existed, a Shopping product group with no products in it, page views and phone calls counted as conversions alongside duplicated purchase actions, a target return set before any data existed, and the store never linked to Google Ads.

In the period before the fixes, the account spent money and recorded zero purchases. Nobody could tell whether Google worked, because it had never been allowed to show it.

What we found

  • Broken product links in live ads
  • Page views and calls counted as conversions
  • Duplicated purchase actions
  • Bidding targets set before any data
  • Merchant Center errors, store not linked

What we fixed

  • Every ad pointing to a live product
  • Purchases as the only conversion
  • Duplicates removed
  • Bidding reset to match the data
  • Integration fixes queued as next steps

What we did

Two accounts, two different jobs

  1. Rebuilt Google around purchases

    Conversions cleaned to purchases only, broken links replaced, missing headlines, descriptions and sitelinks completed, and bidding realigned across Search, Shopping and Performance Max.

    Google AdsCRO & Analytics
  2. Read the Meta quarter honestly

    One automated campaign with creator content did most of the work: 1,032 purchases at a 2.03 return. Women over 45 were the buyers. The cold campaign was switched off.

    MetaAdvantage+
  3. Found the retargeting leak

    Retargeting returned 0.79 because the pixel was not counting real site visitors, so the audience it retargeted was a fraction of the true one. The fix went to the developer.

    Pixel
  4. Planned the next quarter

    User-generated ads from new and existing footage, a second creator collaboration modelled on the winner, a monthly creative refresh, cost caps, and YouTube as a new test.

    Performance Creative

Meta purchases by campaign in the quarter reviewed

Automated campaign with creator content1,032 at 2.03×
All other campaigns, incl. retargeting240
Cold audience test4 · switched off

1,276 purchases and $67.5K in sales in the quarter.

Google turned out to be the cheaper channel

Meta, quarter reviewedGoogle after the fixesCost per purchaseMeta: $29.95Google: $13.87
Meta, quarter reviewedGoogle after the fixesCost per purchaseMeta: $29.95Google: $13.87
Cost per purchase: Meta in the quarter reviewed, against Google in the weeks after the rebuild.

What made the difference

Before you can scale a channel, it has to be allowed to tell the truth. Most of the Google gain came from making the account count what actually mattered.

Every Online Sales level starts with tracking verified in writing, because nothing else can be trusted without it.

See Online Sales