Ad Budget Allocation: How to Split Meta and Google Spend

There is no universal correct split. Put most of your budget where buyers already are: Google if people actively search for what you sell, Meta if you must create demand by showing it to them. Start with a lopsided test split, measure cost per qualified lead in your CRM, and shift money in small steps.

Below: how to pick the starting numbers, how much each platform needs before results mean anything, and how to stop each platform grading its own homework.

Why isn't there one right split for every business?

Because Meta and Google answer different questions. Google Search reaches people who have already decided they need something and typed it into a box. Meta (Facebook and Instagram) reaches people who weren't looking, and persuades them with a picture, a video or an offer.

That difference decides almost everything:

• Intent-led businesses (emergency repairs, legal help, B2B software someone is actively evaluating, "near me" services) win on Google, because the demand already exists and you just need to be there when it surfaces.
• Discovery-led businesses (fashion, wedding photography, home décor, events, most consumer products) win on Meta, because the purchase starts with seeing something you like.
• Most businesses sit in between, which is why a split exists at all.

If you want the full side-by-side of how the two platforms differ, our guide to Meta Ads vs Google Ads for small businesses covers targeting, formats and costs. This article is about the money.

What starting split should you use?

Pick a lopsided split based on how your customers buy, not a polite 50/50. An even split usually means both platforms are underfunded and neither gets enough data to optimise.

These are the starting points we use when planning a new account. They are rules of thumb, not benchmarks, and the first month of data should overrule them.

The split is a hypothesis you test, not a strategy you defend.

How much budget does each platform need before results mean anything?

Each platform needs enough conversions to learn, and splitting a small budget can starve both. If your total budget can't feed one platform properly, run one platform.

On Meta, the delivery system goes through a "learning phase" for each ad set, and Meta's own guidance is that it needs roughly 50 optimisation events (the result you told it to get, such as leads) within a week to stabilise. You don't have to hit that number to get results, but the further below it you sit, the noisier your costs will be.

Google's automated bidding strategies also need steady conversion data. With only a handful of conversions a month, tight keyword lists and simpler bidding usually behave more predictably than fully automated campaign types such as Performance Max.

A quick sanity check you can do yourself:

1. Estimate a realistic cost per lead for the platform (your past data, or a cautious guess).
2. Multiply by the number of leads you need per week for stable delivery.
3. If the answer is more than you can afford on that platform, consolidate.

For example, if you expect around $25 per lead on Meta and want about 50 leads a week in one ad set, that's roughly $1,250 a week on that ad set alone. A business with $1,500 a month total should not be splitting it across two platforms and five campaigns.

What do monthly budget tiers look like in practice?

Below a certain spend, focus beats coverage. Here is how we usually structure accounts at different budget levels (USD, ad spend only, excluding management fees).

In India the same logic applies at much lower absolute numbers, because costs per click and per lead are lower.

What does GRWW's own campaign data say about Meta's role?

In the accounts we manage, Meta was often very cheap for discovery-led and conversational businesses, and clearly weak for a few high-consideration ones. That's why the split depends on the business.

Some real, anonymised examples from Meta accounts we run (lifetime figures, Meta's default 7-day click / 1-day view attribution):

• A clothing brand got 2,564 enquiries at an average of ₹5.43 per conversation.
• A home-improvement (awnings) business got 153 conversations at ₹2.72 each.
• An immigration and study-abroad consultancy got 572 leads plus 525 conversations on about $676 of spend.
• Five wedding photography accounts produced 1,130 enquiries at an average of ₹58.55.

And the honest counterweight:

• An interiors account got just 2 conversations at ₹423 each.
• A removals company paid roughly $13–27 per lead.

We don't have Google data from those same businesses to claim Google would have done better. But high-value, considered purchases, and services people tend to search for when the need arises (moving house is a classic), are where we would test Search before pouring more into social. When Meta's cost per enquiry is low and lead quality holds up, it deserves the bigger share; when it isn't, the answer is rarely "spend more on Meta".

How do you compare the two platforms fairly?

Judge both platforms by the same yardstick, from the same source: qualified leads or sales recorded in your own CRM. Never compare Meta's reported cost per lead against Google's reported cost per conversion and call it a day.

The reasons:

• Different attribution. Meta's default counts conversions within 7 days of a click or 1 day of a view. Google Ads usually uses data-driven attribution for new conversion actions. Both platforms can claim the same sale.
• Different conversion definitions. A Meta "lead" might be an instant form; a Google "conversion" might be a page view if tracking was set up loosely. Check what is being counted. Our step-by-step Google Ads conversion tracking guide helps here.
• Different lead quality. A messaging lead and a search lead behave differently once your team calls them.

Fill this from your CRM each fortnight:

If you're unsure how to read the Meta side, see how to read your Meta Ads Manager report for the metrics that actually matter.

When should you move budget from one platform to the other?

Move money when the cost per qualified lead gap is clear and has held for at least two weeks, and move it in steps of 10–20%. Big swings reset learning and make it impossible to tell what caused what.

Practical rules:

• Wait for volume. A platform with eight leads isn't "winning" over one with six.
• Shift gradually. Moving 15% of budget lets you see whether the cheaper platform keeps its costs as spend rises. Costs often climb as you scale.
• Check for saturation. High impression share on your core Google keywords means extra budget buys little; on Meta, rising frequency with falling click-through rate means the audience is tiring.
• Protect retargeting. Keep a small, steady slice (often 10–20% of Meta spend) for retargeting recent visitors, even if prospecting moves.
• Account for seasonality. A wedding photographer in peak enquiry season and in the off-season should not judge the split on the same month.

Should you split budget by funnel stage instead of by platform?

For accounts over a few thousand dollars a month, yes: give each platform a job, then budget the job. This avoids both platforms fighting for the same people.

A common structure:

• Meta prospecting: reach new people with Reels and image ads.
• Google Search: capture people who search your service or brand name.
• Meta retargeting: remind site visitors and video viewers.
• Brand search on Google: a small, cheap campaign so competitors can't take your name.

Thinking in jobs changes the question from "which platform is better?" to "is each job worth what it costs?", the framing our performance marketing guide to ads, landing pages and tracking is built around.

Frequently asked questions

Is a 50/50 split between Meta and Google ever right?

It can be, once you have data showing both platforms produce qualified leads at similar cost. As a starting point it's usually a mistake, because it halves the data each platform learns from. Start lopsided and let results pull you towards the middle if they justify it.

Should a brand-new business start on Meta or Google?

Ask whether people already search for what you sell. If they do, start with tightly targeted Google Search. If they don't, or if the decision is visual, start with Meta. Either way, set up conversion tracking before spending anything.

How often should I review the budget split?

Every two weeks is a good rhythm for most small and mid-size accounts. Weekly reviews tempt you into reacting to noise; monthly reviews leave money in a losing platform for too long. Make changes in 10–20% steps.

Does Performance Max or Advantage+ change the split?

They change how each platform spends its share, not the logic of the split. Both are automated and work best with good conversion data. Treat them as tests inside each platform's budget rather than as a reason to move money between platforms.

What if my sales happen offline, by phone or in person?

Then platform dashboards undercount both channels. Record the lead source on every enquiry in your CRM, ask customers how they found you, and compare cost per sale from that. Sending offline conversions back to the platforms also improves optimisation.

Get a budget plan built on your own numbers

The right split comes from your sales cycle, your margins and your tracking, not from a template. If you'd like us to review your current accounts and propose a split with clear test rules, ask us for an account review and budget plan or see the kind of performance marketing work we do.