How I warm up Google Ads without wasting money
Search ads, such as Google and Bing, stand apart from other ad platforms. Unlike Meta, where you can see results almost right away, search platforms require a significant investment. Until you feed them enough conversions, the traffic quality will be mediocre at best. It discourages many people from giving it a go. You spend thousands of dollars on high-intent queries and barely get any purchases. The cost per action (CPA) is extremely high, and most quit after a few weeks. Which is a shame, given that search ads can be one of the most profitable and stable sources of revenue.
Lately, with Emery, I faced exactly the same problem. $5k spend, barely any conversions. The exact same funnel that performs well on Meta. The exact same angles. And search traffic is supposed to be high intent. The general consensus says you have to spend at least $10k to get anything approaching quality traffic. And that's if you already know what you're doing – campaign structure, keyword research, cadence. Fine for established businesses, but a deal breaker for a small bootstrapped project.
After some research, it looks like I finally found a way.
Since Google requires real training data and time, the question is how to feed it as cheaply as possible. Solution? Launch a campaign with branded keywords only.
In general, running branded campaigns is controversial. You pay for traffic you’d get anyway. So, overall, I’d be cautious with this approach. But as a warm-up mechanism, it does work, and here’s why:
People searching for your product are the most likely to convert – they either saw your ad somewhere else and are doing research, or someone recommended you.
Branded keywords are much cheaper than competitive category keywords. And even if competitors bid on your branded keywords, the auction is significantly skewed in your favour.
The only downside – many of your existing customers will be searching for the brand too, and you’ll be paying for them as well. However, it’s partially solvable with exclusion lists.
Now, here’s my checklist for pre-warming search ad accounts:
Ensure the tracking is in place – conversions, mid-funnel events, page views firing across all pages. I’ll write another article on this soon, it’s a big topic.
Go to Search Console and find all branded keywords people already use. In some cases, Google already offers a branded/non-branded query filter. If you don’t have it yet, use a regular filter and enter your brand name manually.
Create a search campaign targeting branded keywords. A few rules here:
Location depends on your traffic source. In my case, I’m running Meta ads globally, so I go with worldwide targeting.
Exact match only, despite all the warnings Google shows.
Display Network and Search Partners are off. Otherwise, you’ll get a lot of spam traffic, especially on a new account. Besides, they just don’t make sense for a branded campaign.
Fill in all headlines and descriptions – come up with as many variants as you can.
Don’t set a target CPA yet.
Set the end-of-funnel event as the conversion. Either trial or purchase. Everything else is noise.
Set the budget to get at least 30 conversions a month at your CPA. If you’re just starting out and don’t have enough search volume, you can start with as little as $10 a day.
Set up an exclusion list (Customer Match) to minimise impressions for existing customers. It can be as simple as all your sign-ups.
Once the campaign is live, prepare to be patient and review performance regularly. Google Ads really doesn’t like frequent changes, so we need to minimise two things when scaling the campaign – the size of the changes and how often we make them. My rule of thumb: just a single change a week, be it budget, CPA or anything else.
Here’s what the timeline can look like:
Week 1. Verifying the setup. Checking the search terms to ensure they’re all branded, checking conversion events.
Week 2. Getting the first conversions.
Week 3. After getting a few conversions, I set my initial CPA target. You can use your regular CPA + 20%, or just take the actual value from the previous two weeks.
Weeks 4–8. Start gradually decreasing the CPA. No more than 10% once a week. If we’re to pay for branded traffic, we should pay as little as possible. In general, I’m ready to pay 60–70% of what I pay Meta for cold traffic. So if my regular CPA is $30, I want to gradually get down to $20 while still getting conversions.
After a couple of months of this exercise, you should be able to feed Google enough conversions and start your first non-branded campaign.