2,000% ROI on Google Ads. That's not a typo.

The situation

Melbourne Mint sells gold and silver coins, bullion, and collectibles to investors and enthusiasts across Australia. High average order value. Sophisticated buyer. Good margins. The kind of e-commerce business where one well-targeted Google Ads campaign can change the whole trajectory.

When they came to us, they were spending around $2,000 a month on Google Ads and making back less than they were putting in. Not dramatically. The account wasn't a disaster on the surface. But when you actually dug in, it was spending most of its budget on terms that people use when they're researching, not when they're buying.

The account had been set up by an agency eighteen months earlier. That agency had since stopped returning calls consistently. Nobody had touched the campaign structure in six months.

What was broken

Three things were killing the account.

First, the match types. Almost every keyword in the account was broad match. Broad match in 2024, with Smart Bidding, means Google decides what your ad shows for. Google has a financial incentive to show your ad for things that look vaguely related to your keywords but will never convert. When I pulled the search term report, the account was paying for clicks from people searching "how to invest in gold" (an informational query from someone who doesn't have a credit card out), "gold coin games" (not kidding), and "Melbourne Mint restaurant" (a restaurant that apparently shares the name).

Second, the conversion tracking was wrong. The account was set up to count every page visit to the order confirmation URL as a conversion, but the confirmation page was also accessible from the homepage for logged-in users checking previous orders. The account thought it was converting at 18%. It wasn't. Real purchase rate was around 2.4%.

Third, the campaign structure didn't reflect buyer intent. There was one shopping campaign and one search campaign. No separation between someone searching "buy gold coins Australia" (high intent, ready to purchase) and someone searching "gold coin prices" (low intent, just looking).

What we did

• Fixed conversion tracking first. Installed proper Google Ads conversion tracking via Google Tag Manager, triggered only on purchase confirmation with the order ID passed as a unique identifier to prevent double-counting. Verified against actual order data in the Shopify backend.
• Rebuilt the match type strategy: moved all campaigns to exact and phrase match initially, with a structured negative keyword build process. Paused broad match until conversion data supported Smart Bidding.
• Rebuilt campaign structure around buyer intent: a pure high-intent search campaign for bottom-of-funnel purchase queries, a separate mid-funnel campaign for comparison and research queries (lower bids, different creative), and a shopping campaign with product-level bidding based on margin.
• Added an extensive negative keyword list: 340 terms in the first pass covering informational queries, competitor names, location-adjacent businesses, and every non-commercial modifier we could identify.
• Set Smart Bidding (Target ROAS) only after 30 days of clean conversion data. Running it on wrong conversion data, as the account was doing, actively teaches the algorithm to optimise for the wrong behaviour.
• Rebuilt ad copy around the specific value proposition: competitive pricing, genuine Australian bullion dealer, fast shipping, trusted track record. Called out the things that matter to serious buyers.
• Set up a weekly account review template the owner could use herself: which campaigns to check, what the ROAS threshold was before making changes, and when to call in versus when to leave it alone.

The result

Within three months of the rebuild, the account was running at over 2,000% ROAS. For every dollar spent on ads, the business was making back more than twenty. That's not a number we invented. It's in the Google Ads account, verified against Shopify revenue.

Monthly ad spend went from $2,000 to $3,500 as results justified the increase. Revenue attributable to Google Ads went from roughly break-even to a consistent, predictable, high-return acquisition channel.

The owner now runs the account with a Monday morning review and a monthly strategy call in the Advisory. She understands her ROAS targets, knows when to pause underperforming ad groups, and hasn't needed an agency since the handover.

That was the point.