A busy account that was not paying its way
Easy Tiger sells spirits online into the New Zealand market, with a gin-led range and a loyal repeat following. The store was running Google Ads and the numbers looked active, but the economics underneath were soft. Return on ad spend sat at around 4x, and a large share of the budget was being spent on people who were already going to buy.
Two problems sat behind that. First, the conversion tracking was not clean, so the account was optimising towards signals that did not reflect real sales. Second, the spend was brand-heavy: a meaningful slice of the budget was bidding on Easy Tiger's own name, capturing customers who would have found the site anyway and inflating the headline numbers without adding incremental revenue. With a thin-margin product and an owner who needed every dollar to work harder, that was not a base to scale from.
Fix the tracking, cut the waste, then justify a higher CAC
We began with a full account audit and a tracking rebuild. Before any spend decision could be trusted, the account had to be measuring real purchases and real revenue, so we corrected the conversion setup and tied the reported numbers back to what actually landed in the till. Clean measurement is the foundation everything else stands on.
With trustworthy data in place, we restructured the Google Ads account. We pulled budget away from the brand-heavy waste that was buying existing demand, and reorganised the remaining spend around the product range, so campaigns were segmented by the categories and SKUs that carried genuine new-customer intent rather than lumped under one generic banner. That put the money in front of buyers Easy Tiger would not otherwise have reached.
The decisive piece was financial. Rather than judging a customer on the first order alone, we modelled the value of an acquired customer over 30 and 90 days, building a simple picture of how much a typical buyer was worth once repeat purchases were counted. That LTV-to-CAC view changed the question from "what can we afford to pay for a sale today" to "what is a customer actually worth over the next quarter". Because the modelling showed real repeat value, it justified paying a higher cost per acquisition to win the right buyer, which in turn opened up far more profitable inventory that a first-order-only view would have called too expensive.
Nearly triple the return, and NZ$330k of new revenue
Clean tracking, a product-led account and an LTV-informed bidding strategy reset the economics. Within three months conversions were up 87% and conversion value had doubled, while the cost of acquiring a customer fell. Over the engagement the work generated NZ$330,000 in new revenue and lifted return on ad spend from 4x to 11x.
- NZ$330,000 in new revenue generated over the engagement, on the same store and team.
- ROAS lifted from 4 to 11. Nearly triple the return on every dollar of ad spend.
- Conversions up 87% in three months, with total conversion value doubling over the same window.
- Cost per acquisition down 24% (a fall of NZ$3.67), even as spend moved onto higher-intent, new-customer inventory.
- Brand-heavy waste removed, so the budget bought incremental customers rather than demand the store already owned.
The model justified the move, the data made it safe
Most accounts cannot scale because they are judged on the first sale, which makes good customers look too expensive to win. By modelling the value of an Easy Tiger buyer over 30 and 90 days, we could prove that a higher acquisition cost still paid back, and that turned conservative bidding into confident bidding. Cleaning the tracking made the decision safe, cutting the brand waste freed the budget, and the LTV:CAC view is what allowed the spend to chase real new revenue rather than protect a number. The same financial logic underpins every scaling case we run.