"How much should go to Google vs. Meta vs. TikTok?" There’s no one-size split — but a few principles keep you from burning money.
First separate: intent vs. discovery
Google Search is intent (the user is already looking) — short path to conversion, high ROAS, great for capturing demand. Meta/TikTok lean discovery (creating demand) — strong for prospecting and mindshare. Judging them on the same ROAS yardstick usually penalizes the discovery channels unfairly.
Move by marginal ROAS, not average
What decides "where the next dollar goes" is the marginal return, not the lifetime average. A channel with a great average ROAS may be near saturation, where extra spend returns little.
Test, then shift — in small steps
Don’t move big budget at once. Make small changes, observe, decide; ideally keep a holdout to confirm the *act of reallocating* actually added incremental value rather than coinciding with a market swing.
Don’t set-and-forget
Platform performance shifts weekly (seasonality, auction, creative fatigue). A fixed split drifts off eventually. Either review on a cadence, or hand it to a system that rebalances on the data continuously.