Introduction to Google Search Ad AB Testing

Let’s establish some ground rules before getting started.

First, when we run Google Search Ads, we are looking to capture demand.

This means that we are trying to convert the 3% of the market that is ready to buy.

We are also testing these using Custom Experiments in Google Ads.
‍

Metrics we want to optimize with Google Search Ad AB Tests

There are two metrics we care about with our Google Search Ads AB Tests

• CPL
• CTR

But we aren’t optimizing for both all the time, it actually depends on the ad level that the AB test is being run on.

Tests that are run at the campaign level will be more focused on CPL while tests run at the ad level will be focused on CTR.

More information about this can be found here.

Now we are going to cover 10 AB tests that should be run in every setup Google Ads account to help you optimize your marketing spend.
‍

Campaign Level Tests

At the campaign level, there are a 7 levers that we recommend you pull to see where spend can be more efficiently allocated.

These ab tests will all be focused on driving the most marketing pipeline and lowering the CPL.
‍

Bidding Strategy Tests

Most Google Representatives will tell you to leave your campaign on maximize conversions but this is simply lazy - especially for newer ad accounts that don’t have a lot of conversion data.

A great AB test is to iteratively figure out which bidding strategy works best for your account.

This could be done by first testing maximize conversions with maximize clicks - then the winner of that test can then be tested with enhanced CPC - then the winner of that test can be tested with maximize conversion value.

By doing this, you are iteratively testing different bidding strategies against each other in a controlled environment.
‍

Match Type Tests

Google Ads really wants you to use broad match keywords, but honestly it usually results a large percentage of your ad spend wasted on searches that are not relevant.

A great way to test if broad is right for you is to test phrase and exact match keywords against it.

We would recommend doing broad vs phrase, then the winner against exact.

By doing this you will make a serious improvement in the quality of the people viewing your ads and where your ads are populating in the search.
‍

Location Tests

We have seen really tremendous results when testing hyper specific locations instead of just the US or Canada.

This can be done by using other ad campaign data or your subject matter expertise to understand who to hyper target with geo located ads.

By doing this, you can uncover underserved markets that aren’t getting served hyper specific ads by your competitors.

Ad Schedules

Similar to locations, ad schedules test specific time blocks to get the most out of your ad spend.

A lot of times, your competitors will run out of budget at certain times of the day - and AB testing can help you test if you get better performance at certain times when your competitors run out of budget.

This takes a lot of trial and error but is worth a try if you are starting to see your ads populate higher in Google at certain times of the day.

You can track this natively in Google Ads & analytics.
‍

Audience Targeting Tests

At the campaign level, you are able to target specific audiences in Google Search Ads.

This usually done with observational targeting, however, if you are in a hyper specific target market, then targeting makes perfect sense to apply to your campaign.

AB Tests can be iteratively done here to test in market segments like Company Size 1-250, Company Industry, etc.

When these are layered onto your campaign, they may result in higher marketing pipeline and lower CPL.

Similar to ad schedules, and with all AB testing, this requires a lot of patience but can result in amazing results over time.
‍

Device Type Tests

By default, Google shows your ads on Mobile, Desktop, tablet.

We usually only choose to show ads on Desktop because it has shown us the lowest CPL and highest marketing pipeline, however it is always worth a test.

Inside of the campaign level settings you can adjust your device types to show only on a specific device.

By doing this side by side, you can see which device actually brings in better results.
‍

Advanced Bidding Adjustment Tests

Maybe doing the location or ad scheduling campaigns didn’t result in higher results for your account?

The next best test after iteratively testing those is to try advanced bidding adjustments.

This is effectively the same thing as limiting/promoting ads at specific times or areas but this doesn’t prevent your ads from showing in other areas.

This ab test essentially just promotes or demotes your ads for specific locations or times in an effort to find a combination that drives lower CPL.
‍

Ad Level Tests

There are 2 ad level AB tests that I recommend for all of our clients to try.

Unlike campaign level ab tests where the main metric to test was CPL, ad level tests are focused on CTR.

This is because these changes will primarily influence if someone clicks the ad or not.
‍

Pinned Headline Tests

Using the data on your headlines, you will find some headlines perform best.

Giving a headline a pin could greatly improve your ad’s CTR.

It’s best to iteratively test these over and over to make sure you are always finding the best ad headline combinations.
‍

Keyword Insertion Pinned or Unpinned Test

Using keyword insertion, you are able to place the searcher’s query as a headline in the ad.

We like to test if pinning keyword insertion results in higher CTR or not.
‍

Wrapping Up - How to test campaign level and ad level tests at the same time

When you run ad level tests first, you can simply run campaign level tests after without needing to setup the ad level tests each time.

This helps keep your account organized and allows for you to always be testing two things at once.

Double the learnings = double the earnings.

Until next time