-
Yeah this is pretty common with big messy catalogs like that, especially in smaller CEE markets. PMax tends to work better when it has more data to play with, so splitting too early just starves each segment.
At that AOV and budget, you’re kind of forced into a volume game, which is why the “one big feed + prune losers” approach is holding up. Not ideal, but it makes sense.
One thing that’s helped me in similar setups is not splitting campaigns, but splitting the *feed logic*. Like using custom labels to separate margin tiers, price bands, or even “proven vs unproven” products, then adjusting targets or exclusions gradually instead of hard campaign splits. Keeps the data pooled but gives you some control.
Also worth checking if your best products are actually losing because of competition or just getting drowned in the catalog. Sometimes isolating only the absolute top performers into their own campaign with a protected budget works, but only if they already have consistent conversions.
Honestly though, with 10k SKUs, €20 AOV, and €100/day, there’s just a ceiling. Sounds like you’ve already found the most stable version and now it’s more about squeezing efficiency than expecting a big unlock.
