Media buyer notes: what small budgets get wrong on Meta and Google

Yusuf's ex-colleague, media buyer, Izmir. The five small-budget sins, with fixes:

One campaign per test, not five ad sets of 4 euros. Consolidate or the algorithm learns nothing. Minimum 15-20 euros a day per test cell or do not test. Broad beats narrow now. Creative is the targeting. Three angles, two formats, refresh weekly. Narrow interests plus tired creative is the classic burn. Search first for capture. Branded + competitor + high-intent, always on, boring, profitable. It funds the Meta experiments. Track server-side. Browser pixels miss a third of conversions at small scale and you optimize on lies. One afternoon of setup, permanent truth. Kill rules written before launch. 2x CAC, zero sales, dead. Written down, not felt. Feelings scale budgets. Rules scale profits.

Ex-colleague approved. The kill-rules-written-down point is the one I steal for every client call. Rules before launch, because after launch everyone believes unconditionally in their ugly baby creative.

Elif trained me, confirm all five. The server-side tracking afternoon is the highest-ROI afternoon in ads. Optimizing on missed conversions is driving with a painted windshield.

Cohort wall grows: add "optimize on truth". Server-side tracking plus cohort retention is the full picture. Clicks are vanity, cohorts are sanity.

E-commerce operator view: the search-first point funds everything. Retourly's branded search prints the money Meta experiments spend. Boring profits, exciting tests.