A decade ago, it was very common to build out dozens or hundreds of Google Ads search campaigns in a single account. This used to be an acceptable organizational system when we were all micromanaging simple CPC keyword bids. But that’s not the case any longer. In the era of automated bidding, having too many campaigns is a performance killer.
If you’re using a modern value-based bidding strategy to optimize for maximum conversion value at a given budget or to flex your budgets up and down so long as they hit a target ROAS, you need the fewest number of campaigns possible.
Why is this? It’s because the machine-learning systems that power these smart bidding strategies need to be trained on as much data as possible to learn what works and what doesn’t. If you fragment your account into too many campaigns, each campaign has fewer conversions in it, and will improve more slowly. If you have far too few conversions, your campaign may never get off the ground.
Using Brand Keywords to Speed Up Learning
Because of this fundamental shift in how campaigns are optimized, much of what we all learned as “best practices” from 5 or 10 years ago is simply wrong now. For example, many people think you should automatically split out “brand” keywords into a separate Brand campaign in every new account. This wasn’t actually necessary back in the day, but it also didn’t cause any problems. Now, it is probably better to lump your brand keywords into the same search ads campaign as everything else when you’re first getting started. People searching for your brand terms provide valuable training for your new campaign. Lump them in there with all your other ad groups when you’re starting out.
Yes, Brand keywords in your campaign will skew results towards that low-hanging fruit. But it will help the other ad groups in the campaign train faster on things like what times of day people buy, and what days of the week. It will help the other ad groups learn what audiences convert. Later on, when you have a larger campaign with a lot of conversions, you can then split the brand keywords out into a separate campaign so you can treat them differently.
What Size Should Campaigns Be?
That begs the question, “When can I split out those brand keywords?” The quick answer is, not until the non-brand ad groups are getting 15 to 30 conversions per month. But let’s explore the reason why…
When we’re using value-based bidding like Maximize Conversion Value (MCV) or Target ROAS (tROAS) strategies, Google has different recommendations on how large campaigns should be. For MCV, there is no official guidance. But for tROAS, Google explicitly says that campaigns generally need 15 conversions per month in order to work well. Dip below this level and performance suffers. Since many businesses have seasonality in their results, we really don’t want to allow a campaign to have fewer than 15 conversions per month in a slow month, which means such a campaign would have more conversions than that in a good month. A good rule of thumb is that we want every campaign to fluctuate around a range of conversions per month that never dips below 15, so in practice this often manifests as campaigns with 15 to 30 conversions per month.
It’s useful to keep this minimum level for Maximize Conversion Value campaigns as well, since they are actually often flipped over to tROAS strategies as business goals shift and change over time. If we ensure no campaign is ever under 15 conversions per month, we have maximum flexibility to adapt to changing business goals while still having campaigns that can perform well.
Never Split Up a Campaign With Fewer Than 30 Conversions per Month
If you never want a campaign that has fewer than 15 conversions per month, that means that the smallest a campaign could possibly be before you consider splitting it is 30 conversions per month. Going back to the previous example, if you’re starting out with a small account, it doesn’t make sense to split out brand keywords into a separate campaign until you’re over that threshold.
In reality, you might have 20 brand conversions and 10 generic keyword conversions. That’s 30. But you still shouldn’t split that campaign to treat those keywords differently until the non-brand keyword conversions pass that 15 conversion mark, and preferably by a significant margin. So maybe you wouldn’t actually split that campaign up until you were at 40 or 50 conversions per month.
What About Performance Max Campaigns?
This is a bit fuzzy. The PMax documentation in Google Ads Help isn’t very clear, but in various presentations from Google, we have seen recommendations that PMax campaigns should have 20-30 conversions per month minimum, up to 50 conversions per month minimum.
What’s the right answer? There probably isn’t one precise number that is always true. But the more conversions in a campaign in a given period of time, the more data it has to optimize, and the higher the performance will be, as a general rule.
You Need Fewer Campaigns With More Conversion Data
The bottom line is that you do not want a lot of small campaigns with low conversions. You want as few campaigns as possible that cluster the greatest volume of conversion data possible. When you are thinking about your accounts, instead of thinking about how you want to break out more campaigns to micromanage budgets and reporting, you should be thinking about how you can combine campaigns together to achieve greater performance.
In the age of machine learning, data quality is what leads to good financial performance. And these structural decisions have a huge impact on campaign performance over time.
If you think that the ad agency you’re working with right now just doesn’t get it, reach out and chat with us. Maybe we can help.