Advantage+ Sales Campaigns for Digital-Goods Checkout

An Advantage+ sales campaign needs roughly fifty optimisation events a week to get out of learning, and value optimisation needs somewhere near a hundred purchases in a rolling seven days before the bidder's revenue predictions are worth reading. Below that, this engagement is the wrong purchase and we will say so on the first call rather than after the invoice. What we run these on: Digital goods with an instrumentable checkout. Seat upgrades and plan changes for developer tools, course and cohort enrolments, game purchases and downloadable content, subscription tiers. We do not run apparel, furniture or general retail catalogues. Returns modelling, size variants and stock-driven feed churn are a genuinely different craft, and there are shops in Amsterdam who do it better than we would. Structure, and how little of it there is: One campaign per market with real budget behind it, rather than nine campaigns built to make a slide look busy. The existing-customer budget cap is set from your own repeat-rate data, not left on a default that assumes a retail buying cycle; for a tool with annual contracts and for a game selling DLC every six weeks, that number is nowhere near the same. Cost cap goes on only when you have a defensible marginal figure to cap against, which usually means gross margin after payment fees and support cost, not blended CAC. Creative slots are filled to the count the campaign can actually cycle through, and no further. What we do not do inside the campaign is the part buyers find hardest to accept: no audience splitting, no manual placement exclusions on a hunch, no bid nudges. Advantage+ takes those decisions away and does them better. Paying somebody to fight the bidder is paying for the privilege of losing slowly. Measurement: Seven-day click and one-day view is the reporting standard, and it is not the number we argue from. We run geo holdouts with Meta's open-source GeoLift package in R, using the Netherlands and Belgium as separable cells where the language split allows, or regional cells within a larger market where it does not. A readable test needs four to six weeks and a spend delta large enough to clear noise, which means we plan the test before the quarter, not after somebody questions a result. Creative: briefed, not produced: We write the brief, set the concept count per cycle, specify hook variants and the format mix, and review what comes back against what the account is short of. Production sits with your studio or a partner. We are not a creative agency and taking a margin on other people's editing would make our creative feedback worth less. What we refuse: Daily bid adjustments presented as a deliverable. Standing weekly calls whose function is to demonstrate effort. Presenting Ads Manager ROAS to your board without a holdout result standing next to it, which is the single most expensive habit in this channel and the reason budgets get cut in the quarter after they get raised. Who this is not for: Accounts below about forty thousand euros a month in working media, because the maths at the top of this page does not bend. Businesses whose checkout runs on a third-party page they cannot instrument, where we would be optimising toward a purchase value we are guessing at. And anyone who needs the answer to be yes before the test runs.