“Should we be on Google or Meta?” has never had a general answer, and the version of it people usually argue about (which platform performs better) isn’t the question that decides it.

The question that decides it is whether people already go looking for the thing you sell. Everything else is detail, and most of the detail only matters once you’ve settled that part.

The difference everything else follows from

Google Ads sells you access to a question. Someone typed “emergency locksmith parramatta” and you get to be in front of them while they’re typing it. You’re buying demand that already exists, at whatever price the other people chasing that same demand have bid it to.

Meta sells you access to attention. Nobody opens Instagram intending to buy anything. You interrupt them with something good enough to stop the scroll, and when it works you’ve created a customer who wasn’t looking for you an hour ago.

Capture versus create. Both are legitimate businesses to be in. They suit different companies, and the most common mistake I see in Australian accounts is choosing on the basis of which platform the founder personally uses.

Check whether anyone is searching first

Open Google’s Keyword Planner, type in the words a customer would use for what you sell, filter to Australia, and look at the volume. Ten minutes, free, and it settles most of the argument.

If there are thousands of searches a month, start with Google. Someone typing your category name is further along than anyone you can reach on Meta, and your ad gets to answer a question rather than pick a fight for attention.

If the volume is close to nothing, that’s just as useful to know. Either your category has a name customers don’t use, or the problem you solve is one they don’t know they have. Both are Meta problems. You can’t buy demand that isn’t there, and pouring search budget into an empty auction is a slow and expensive way to learn it.

Two honest exceptions. Genuinely new categories often show no volume and later show plenty, so a nil result on a new idea is not a verdict. And plenty of impulse products carry real search volume but still do better on Meta, because the search happens after somebody has seen the thing.

What Google is genuinely better at

Urgency first. A cracked windscreen is a today problem, and nobody solves a today problem by scrolling.

Novus Glass is the example I keep using. When we took over the account the ads were running and not working: a click-through rate of 0.82%, a conversion rate of 1.94%, and every conversion costing $58.48. The fix wasn’t more budget. It was tighter keywords, ad copy written for a person with a broken windscreen instead of for the brand, geo-targeting and scheduling that stopped paying for the wrong people at the wrong hours, and landing pages that matched what the ad promised. Click-through rate more than doubled, conversion rate reached 7.27%, and cost per conversion fell 52% to $27.54 while total spend came down a quarter. The Novus Glass case study has the full sequence.

Look at the shape of that result. The demand was already there. The account was simply failing to collect it, which is the most common thing wrong with a search account and the least glamorous to fix.

Google also wins where the purchase is considered and researched, where buyers compare named suppliers, and anywhere the product is hard to photograph. B2B services with a long sales cycle rarely stop a scroll and rarely need to.

What Meta is genuinely better at

Anything visual, anything priced low enough not to need a meeting, and anything people didn’t know existed.

Meta’s real advantage is cheap reach and fast creative feedback. You can put four versions of an idea in front of a lot of people quickly and find out which one anybody cares about, which is worth having even if you never spend another dollar there. It’s also the better place to reach people by who they are rather than by what they typed, which is the only option in a category nobody searches for.

And it’s where you go back to people who already visited. Retargeting on Meta is usually the cheapest conversion in the whole account, which is also why it flatters itself in reporting. More on that shortly.

The tax is creative. Search ads are text and they survive being average. Meta ads live or die on the asset, and an ad that worked in March will be tired by June whether or not anything else changed. If nobody in your business can produce new creative every few weeks, budget for someone who can, or plan for performance to decay on a schedule.

The cost per click comparison is a trap

Meta clicks are cheaper than Google clicks in nearly every category. True, and close to useless.

A click from someone typing “buy X near me” and a click from someone who tapped an image out of idle curiosity are not the same asset, and comparing their prices tells you nothing about which one made you money.

The comparison worth making is cost per customer, and then whether that number sits under what a customer is actually worth to you. Take the gross margin on one customer, decide how much of it you’re willing to hand over to buy one, and you have a ceiling both platforms have to come in under. Neither of them will tell you that number, because neither of them knows it.

The two platforms will never agree

Run both and you’ll be shown more conversions than you made sales. Nobody is lying to you. Each platform counts conversions it believes it influenced, on its own attribution window, and Meta counts views as well as clicks while Google mostly counts clicks. Add the two together and you’ve counted every customer they both touched twice.

Decide on one source of truth before you start, ideally your own GA4 property or the CRM where deals get marked won, and judge the channels against that. Use the in-platform numbers for the job they’re good at, which is deciding what to keep and what to kill inside that platform.

How the delivery systems arrive at those decisions in the first place is a longer story, and it’s in behind digital ad delivery.

If you can only fund one

Splitting a small budget across two platforms usually produces two accounts that never learn anything. Both run on automated bidding that needs a steady flow of conversions to work with, and half a trickle teaches a machine less than a whole one.

Fund one properly instead. If the search volume exists and you sell something people go looking for, that’s Google. If the volume isn’t there, or the product has to be seen to be understood, that’s Meta. Get it working, then add the second channel with new money rather than by halving the first.

Running both, and in what order

Once there’s enough budget for two they stop competing. Meta creates demand and the branded searches that follow it; Google collects them along with everything else. Businesses running both well tend to watch branded search volume rise while the Meta account is switched on, which is the two channels doing their actual jobs.

So sequence it that way. Capture what already exists, prove the unit economics stand up, then spend on creating more demand. Doing it in reverse means paying to build interest you have no efficient way to collect.

If you’d rather have this conversation about your business than about advertising in general, that’s where our advertising work starts, and the search half of it lives in paid search.

Google Ads or Meta Ads: where your budget belongs