$221k spend, 3.34x ROAS against a 2x target, 5 months from zero
An Australian ecommerce brand selling anti theft and tracking products for cars and driveways. They had a Meta channel that was already working, paid traffic running profitably, ad creative tested, customer base built. Google Ads was untouched territory. No campaign history, no conversion data, no quality scores, no audience signals, nothing for the algorithms to work with. The brief was simple. Prove Google could work too. Hit a 2x profit ROAS or shut it down. Budget was scalable from day one if I could keep it positive.
Five months in: $737K revenue from $221K spend at 3.34x ROAS, against a 2x target. The dip in Feb is the PMax learning phase, the recovery is the algorithm settling
Don't pick one channel and pray, layer Search, Shopping and PMax from the start so each one feeds the next. Option B: Treat Google like Meta, buil
GA4, Google Tag Manager, transaction value imports, all wired and tested before the first ad ran. Without a clean revenue signal feeding back into the bidding, none of the ROAS optimisation that came later would have held up. Most accounts skip this step. I won't.
Started with brand Search to capture the demand Meta had already built, then rolled out non brand Search around the highest intent product terms (Driveway Defender, Car GPS Tracker, Road Guardian). Brand Search hit 2.87x ROAS, non brand product Search hit 3.86x. That gave the account its first wave of conversion data.
Built the Merchant Center feed cleanly, fixed the product titles, descriptions and image quality before a single Shopping click was bought. Launched the All Products Shopping campaign feeding the same conversion event. Shopping settled at 2.65x ROAS, comfortably above target.
Want this for your account?
Hire on Upwork