Meta vs Google for eCommerce: where should the next €10,000 go?
The two channels solve different problems. A decision framework based on demand volume, margin and creative capacity — plus what changes once you are spending on both.
Short answer
Send the next €10,000 to Google if people already search for what you sell and non-brand search impression share is below roughly 60%. Send it to Meta if demand must be created, your product is visually demonstrable, and you can ship four or more new creative concepts per month.
Key takeaways
- Google harvests existing demand; Meta creates it. Most brands need both, in that order of certainty.
- Check non-brand search impression share before adding Meta budget — unclaimed search is the cheapest revenue in the account.
- Meta scales further but demands creative volume; without it the budget just raises frequency.
- Shopping/Performance Max and Meta advantage campaigns overlap; measure the pair blended, not separately.
- A 60/40 split in either direction is common; 50/50 by default is usually a sign nobody decided.
Question one: does the demand already exist?
Open Google Ads and check search impression share for your non-brand terms. If you are capturing under 60% of available impressions on commercially relevant keywords, there is revenue sitting there at a known intent level. Buy that before you fund demand creation.
If search volume for your category is thin — a new product form, an unfamiliar solution, a design-led purchase — then Google has little to harvest and Meta is where the growth lives.
Question two: can you feed Meta?
Meta budget converts into performance through creative. A brand that can produce eight concepts a month will scale on Meta; a brand that can produce one will watch frequency climb and CPA drift upward within three weeks.
Be honest about capacity before allocating. If creative production is the constraint, the highest-return use of the next €10,000 may be €7,000 of media and €3,000 of creative rather than €10,000 of media.
| Situation | Send it to |
|---|---|
| Non-brand search impression share under 60% | |
| Category has little search volume | Meta |
| Product is visually demonstrable | Meta |
| Margin under 35% and CPCs high | Google (brand + shopping first) |
| Creative pipeline is empty | Google, then fix creative |
| Retargeting audience is small | Meta prospecting |
What each channel is actually good at
Google converts intent efficiently and predictably. Shopping and Performance Max put products in front of people mid-decision, and brand search protects the demand your other channels create. Its ceiling is set by how many people search.
Meta creates demand and has effectively no volume ceiling, but performance is more volatile, more creative-dependent and more sensitive to signal quality. It also feeds Google: brands that scale Meta usually see brand search volume rise within weeks, which quietly improves blended Google performance.
- Google: certainty, efficiency, limited ceiling
- Meta: volume, discovery, creative-dependent
- Meta lifts Google brand search — measure the pair together
- Neither channel should be judged on last-click alone
The halo effect and why splits look wrong
When Meta spend rises, Google brand search impressions rise and Google's reported ROAS improves — on demand Meta generated. Cut Meta and Google's numbers look strong for a fortnight before total revenue falls. This is the single most common cause of a bad channel decision.
The remedy is a blended view plus, once a year, a genuine holdout: pause one channel in comparable regions and measure total revenue rather than platform ROAS.
A practical allocation for a €50,000/month brand
A typical mature split at this level is 45–55% Meta prospecting and retargeting, 30–40% Google brand plus shopping and non-brand search, and 10–15% in a third channel being tested against a clear break-even ROAS.
Review the split quarterly against marginal contribution profit, not monthly against ROAS. Monthly reallocation resets learning on both platforms and costs more than the imbalance it corrects.
FAQ
Should I pause Meta if Google shows a better ROAS?
Not on that evidence alone. Google's ROAS is inflated by brand searches that Meta generated. Run a geo holdout or a phased pause and measure total store revenue before making the call.
Is Performance Max enough on its own for Google?
Rarely. Performance Max blends brand and non-brand traffic and can flatter itself with brand searches. Keep a dedicated brand campaign and, where volume allows, standard shopping or search campaigns so you can see non-brand performance separately.
What about TikTok in this comparison?
TikTok is a Meta-style demand-creation channel with a younger audience and a stronger requirement for native creative. Add it only once Meta is funded above its optimisation floor and creative production is comfortably keeping up.
Keep reading
Want this run on your own account?
Request a free ad account audit. We review tracking, account structure, creative mix and wasted spend, and send back a prioritised 90-day plan within 48–72 hours of getting access.