The plateau is structural, not algorithmic

Almost every account we inherit has hit a ceiling in Advantage+ Shopping somewhere between forty and eighty thousand dollars in monthly spend. The instinct is to blame the algorithm or the auction. The reality is that the structure underneath the campaign was built for the spend level it is at today, not the one the brand is trying to reach.

Creative volume is the real lever

The single biggest predictor of whether an Advantage+ campaign keeps compounding past the plateau is the rate of new creative entering the campaign each week. Not iterations of the same concept — genuinely new concepts. We hold our clients to a minimum of eight new concepts per week with two to three variants of each, and the accounts that scale are the ones that hit that number consistently for a quarter.

Existing customer caps and value rules

The default existing-customer cap in Advantage+ Shopping is rarely the right number for a brand with a real repeat-purchase tail. We tune it based on the actual share of revenue that comes from repeat customers in the trailing ninety days, and we revisit it monthly.

Value rules are where most accounts quietly leave money on the table. A correctly configured value rule lets ASC bid more for the customer cohorts the business actually wants — high-AOV first-time buyers, customers from priority geos, segments with known higher LTV. Without value rules the system optimises for the cheapest conversion, which is almost never the most profitable one.

Account architecture

We run a single Advantage+ Shopping campaign per geographic cluster, never per product line. Splitting by product fragments learning and starves each campaign of the volume it needs to exit the learning phase. Product-level performance is read in reporting, not enforced through campaign structure.

Alongside ASC we run a small set of supporting campaigns: a top-of-funnel video prospecting campaign that exists only to feed the retargeting pool, and a manual retargeting campaign for the moments when ASC under-allocates to warm audiences. These are sized at roughly fifteen percent of total spend combined.

The pixel and the signal

None of this works on a noisy pixel. Server-side events through the Conversions API, with proper deduplication against the browser pixel, are non-negotiable. We also push offline events for refunds and cancellations so the system learns on net revenue rather than gross.

How the rebuild compounds

The first two weeks after a rebuild typically look flat. The campaign re-enters learning, frequency settles, and the new creative library starts to find its winners. By week four we usually see the previous spend ceiling break, and by week eight the account is at a new equilibrium thirty to fifty percent above the old plateau. The pattern is consistent enough that we now use it as the default playbook for every Meta-led account that has stalled at scale.