The debate gets framed as a platform war, but it is really a question about your customer. Answer it honestly and the budget allocates itself.
The deciding question
When your ideal customer’s problem occurs, do they search for the solution? A blocked drain, an emergency dentist, “seo agency london”: these are searched with intent, and Google Search puts you in front of that intent at its exact moment. Nothing on Meta replicates that.
But if your product is one people do not know to search for, a new gadget, a subscription box, a better version of something they tolerate, search demand barely exists to capture. Meta’s job is creating demand: interrupting the scroll with something compelling enough to want. Different mechanism, different economics.
What the numbers look like when each is right
Run on the correct side of that line, both platforms produce unglamorous, excellent economics. I have held Meta cost per messaging result between $0.21 and $0.83 for an e-commerce brand across parallel lead, sales and awareness campaigns. On Google’s side, an app install campaign delivered 3,827 installs at $0.12 each on $214 of spend. Neither number came from the platform being magic; both came from the platform matching the demand type.
The practical starting rule
Search demand exists for your offer: start with Google Search, add Meta remarketing once traffic flows, expand to Meta prospecting when search demand is fully captured. Search demand barely exists: start with Meta prospecting, and run a small Google brand campaign so the people Meta convinces can find you when they search your name, because they will.
The mistake that beats both platforms
Splitting a small budget across both “to test”, too thin on each to buy meaningful data, then concluding neither works. Concentrate first. Win one channel, then diversify from a position of knowledge rather than hope. And before any of it, verify your conversion tracking, because optimising against miscounted results is just automating the waste.