All articles
Budget Intelligence6 min read

How many ads per ad group in Google Ads?

How many ads per ad group in Google Ads comes down to your weekly conversions divided by every campaign, ad group and ad you built.

The account arrives looking organised. Four campaigns, 40 ad groups, 3 ads in each, keywords sorted into themes by somebody who cared. Then you open the conversions column for the month: 11 leads. Not 11 per ad group. Eleven, across all of it.

The meeting that follows is always about the ads. A fourth one, better headlines, the rotation setting. I have argued that side of it, which is how I know how comfortable it is.

Short answer: The number of ads per ad group in Google Ads is set by conversion volume rather than by a rule of thumb. Divide your weekly conversions by every campaign, ad group and ad you created. When that division leaves each ad holding a handful of events, the account has too many containers, not too few ads.

The takeaways

  • The ad count is the last division in a chain. Four campaigns holding 40 ad groups with 3 ads each is 120 slots. Spread 11 weekly conversions across 120 slots and no row in the report means anything.
  • Google guards the cheap split and leaves the expensive one wide open. There is a hard ceiling on how many ads run at once inside one ad group, and effectively none on how many ad groups you create.
  • Consolidation buys readable numbers and nothing else. Merging 40 ad groups into 6 adds zero conversions. It moves the same 11 events into fewer containers so a difference between them can be seen.

So how many ads per ad group should you run?

Start with 2, and add a third only when that ad group produces enough conversions in a week for both of the first two to finish with a number you would defend in front of a client. On plenty of accounts that week never arrives, so the answer stays at 2 and the interesting question sits somewhere else.

The advice in circulation is 3 to 5, quoted per ad group regardless of the account it sits in. Run the division above: 11 conversions a week, 120 ad slots, each ad asked to earn a decision out of roughly a tenth of a conversion. No rotation setting repairs that.

There is also less to win here than the search volume suggests. A responsive search ad already assembles its own headline and description combinations, so a second ad only earns its slot by carrying a different message.

Why is the ad-group count the more expensive decision?

Because the bid strategy lives on the campaign while your read on an ad lives at ad level, so an extra ad group buys control at neither end and divides your conversions on the way through. The split feels like precision when you build it. What it does is thin every row underneath.

The habit is inherited. Narrow ad groups made sense when you set a bid per keyword and needed the ad text welded to the query. Smart Bidding moved that decision up a level, nobody went back and unbuilt the structure, and accounts kept the fragmentation without the reason for it.

Nobody searches for how many ad groups they should have while building the fortieth one. It gets added mid-task, for a theme that felt distinct on a Tuesday, and no screen in Google Ads objects. The cost lands months later.

What does the limit on ads inside an ad group tell you?

Read the asymmetry. Google caps how many ads can run at once in a single ad group and puts effectively no practical ceiling on how many ad groups you create. The platform guards the division that costs least and hands you the one that costs most with no guardrail at all.

That cap explains why this question is hard to get badly wrong: inside one ad group you cannot divide the traffic more than a few ways. You can build 200 ad groups on a 40 euro daily budget in an afternoon and nothing warns you.

So the question people search is the safe one. The decision that ruins the account sits immediately above it, made without a rule of thumb and without anybody counting the conversions available.

When is merging ad groups the biggest change available?

When the account has more containers than weekly conversions. That is the whole test and it takes a minute: count your enabled ad groups, then count conversions in the last 7 days. If the first number is larger, no amount of ad-level work produces a readable result, and merging is the most valuable edit on the account.

Merge by the decision rather than by the theme. Two ad groups belong apart when you would spend differently on them: a different landing page, a different margin, a geography you bid separately. Two that exist because the keywords sounded like different topics belong together.

Do it in one pass rather than in weekly nibbles. A restructure this size resets what the bidding has learned, and the Google Ads learning phase is expensive enough to pay for once.

What does consolidating not fix?

Consolidation rearranges your conversions. That is all. Merge 40 ad groups into 6 and you still have 11 conversions a week, thin ground for judging anything, and a campaign that was losing money goes on losing money in a tidier layout.

I want to be plain, because a restructure is satisfying work and it flatters everyone in the room. You get numbers you can read. You do not get demand. If the account was short of conversions because the offer or the market is the problem, a cleaner structure only makes that legible sooner.

What consolidation does buy is the right to ask one question at a time. Six ad groups sharing 11 weekly conversions cannot answer six questions. They can answer one, and that is the gap between an account that learns something each month and one that gets rebuilt every quarter.

Deciding the shape before you fill it

Structure is a budget decision wearing an operational costume, which is why the useful version of this argument happens before anybody opens the campaign builder.

That ordering is what budget intelligence is built around here: 3 to 5 prioritised campaign plans drawn from 12 weeks of your real performance, each carrying its own conversion goal. Adscalr does not touch your Google Ads structure, bids or rotation settings. The Meta version of the count question runs on the same logic: how many ads per ad set.

Count your containers before you count your ads. The answer to the second question was hiding in the first one.

This is the thinking behind Adscalr.

See the product