Core carries most of the budget by default. Bestsellers convert on proven terms, so the dollar works hardest there.

The split moves off core as the auction tightens, as ads already carry more of your sales, and as the target climbs. Each is a sign the next core dollar returns less, so spend shifts to growth and new launches where there is still room to gain.

The small launch share is exploration, not a growth bet. It funds the testing that finds your next core product.

Some situations set the split outright. A launch funds launches. A clearance funds end-of-life stock. Everything else follows your core's headroom.

If you tell us how many products sit in each stage, Balanced sets each stage's budget share toward its share of products, the fairest starting point when you're funding a subset of the catalog rather than the whole account, and the share-of-catalog slider concentrates or spreads that from there. From there, estimated ACoS still does its job: a stage that converts more efficiently can earn more than its product count alone would suggest, so nudge the sliders once you see the numbers.

Estimated ACoS itself comes from the same products-to-revenue math, not a flat guess per stage. Bestsellers typically run about 20% of products and 80% of revenue, so by default core's product count carries far more revenue per item than the rest of the catalog, and that gap is what makes its estimated ACoS run well under break even. Growth, launches, and declining split the remaining revenue in proportion to their own product counts, then two real-world adjustments apply on top: clearance is priced at break even, since selling at cost means little to no margin left over, and a launch carries an extra premium beyond that, since a product with no sales rank yet is rarely profitable from day one. Type your own product counts above to sharpen all four estimates at once.