Measure Advantage+ shopping campaigns, now called Advantage+ sales campaigns, on new-customer revenue from a source outside Meta, and confirm the result with a holdout test before you scale. Meta’s reported ROAS for these campaigns usually includes a lot of existing customers, so on its own it overstates what the campaign added.
Why does Advantage+ ROAS look better than my other campaigns?
Advantage+ campaigns let Meta decide who sees your ads, and the algorithm is very good at finding people who were already likely to buy. That often means past customers, recent site visitors and people on your email list. Those buyers convert at high rates, so reported ROAS climbs, but many of those orders would have happened without the ad. The same effect shows up in manual retargeting, which I wrote about in is retargeting ROAS real.
What happened to the existing customer budget cap?
Meta folded Advantage+ shopping into the broader Advantage+ sales campaign type, and during that transition the old setting that capped the share of budget spent on existing customers went away. Ad sets came back instead, so the usual workaround is to separate audiences across ad sets and put spending limits on the one aimed at existing customers. Meta changes these controls often, so check what your account shows today before you rely on any guide, including this one.
How do I separate new-customer revenue from returning?
Start by defining existing customers in your Meta ad account settings, usually by uploading your customer list and connecting your purchasers audience. That improves Meta’s own reporting, but it is still Meta grading its own work.
The better check comes from your store data. Pull new-customer orders and revenue for the campaign from an attribution tool that knows who has bought before, and calculate new-customer ROAS from that. ThoughtMetric reports new and returning customer revenue and ROAS at the campaign, ad set and ad level, so you can see how much of an Advantage+ campaign’s revenue came from people who had never ordered. (Disclosure: ThoughtMetric sponsors this site.) If a campaign reports a 5x ROAS and only a third of its revenue is new customers, its new-customer ROAS is closer to 1.7x, which is a very different budget conversation.
Should I trust Meta’s numbers or my attribution tool?
Neither one alone. Meta sees view-through conversions and in-app behavior your tool cannot see. Your attribution tool sees the full order history and every other channel the customer touched. When the two disagree, look at the direction and the size of the gap over several weeks rather than any single day. A gap that widens as you scale usually means Meta is crediting itself for orders other channels drove. The reasons behind that gap are in why Meta ROAS is inaccurate and is Meta-reported ROAS accurate.
How do I test whether Advantage+ is incremental?
Run a holdout. The simplest version is a geographic split, where you turn the campaign off in a set of comparable regions for two to four weeks and compare total store revenue between the regions that kept it and the ones that did not. Meta also offers conversion lift studies to some advertisers, which is worth asking your rep about.
Keep everything else stable during the test. If you launch a sale or change your email schedule halfway through, the result is noise.
How long should Advantage+ run before I judge it?
Give it at least two to three weeks at a steady budget before drawing conclusions, longer if your order volume is low. The campaign needs time to exit its learning phase, and your attribution tool needs enough orders to show a stable split between new and returning customers. Judging it after four days of high returning-customer revenue will make it look better than it is, and judging it during a sale will make everything look better than it is.
Should I exclude existing customers entirely?
Not always. Some retargeting of recent visitors and lapsed customers is worth paying for. The problem is when the campaign spends a large share of budget on people who were going to buy this week anyway. A practical approach is to keep one ad set aimed at prospecting with your customer list excluded, and a separate, smaller ad set for existing customers with a spending limit. Then measure each on its own terms, new-customer ROAS for prospecting and incremental revenue from a holdout for the returning-customer ad set.
What should I track week to week?
- Share of campaign revenue from new customers, from your store data rather than Meta.
- New-customer ROAS by campaign, compared against your target CAC.
- Blended MER, so a campaign that looks great inside Meta has to show up in total revenue too.
- Spend on the existing-customer ad set, if you split audiences, against what you intended.
Leave a Reply