The Importance of A/B Testing in Google Ads

Google ads remains one of the most effective forms of advertising for businesses looking to increase their online presence, sales and success. Over 80% of internet users choose Google as their search engine of choice, and with so many competitors out there, Google Ads are a great way of ensuring your business is the first thing a user sees when searching for a specific term.

But just because you’re running a Google Ad, doesn’t mean that business will grow overnight. With so many other businesses utilising ads, there’s a competitive landscape out there. Fortunately, there is a way to ensure your adverts are as efficient as possible, ensuring they appear to the people, for less costs. The way to do this is by A/B testing.

What is A/B testing?

Put simply, A/B testing is when you test two different ideas, to see which one performs better. You are comparing Example A, with Example B and measuring the results. A is often considered to be the “control” group; in this case, the advert you have been running all this time. B is the variant, a version of A that has a slight change to it for you to measure.

When A/B testing is completed, you measure the results, and see which performed best. The better performer can then become ‘A’ when you test again.

How long should I A/B test for?

Google suggests running an A/B test for a minimum of 2 weeks, however, in some cases you may see conclusive results after just 7 days.

For definitive results, it is recommended to A/B test for at least 1 month, to allow Google to serve your ads to the right audience, and to gather more data from the test.

Remember: the longer you run your test, the more accurate and conclusive the results will be.

What can I A/B test with Google Ads?

You can test almost anything with google ads, depending on the kind of business you run. The more you test, the better understanding you will have of what adverts work for your customers.

Different aspects you could test include:

• Headlines – this the brief text that appears at the top of your Google ads.
• Descriptions – the longer text that appears in the body copy of your google ads.
• Keywords – the actual search terms you want people to use to find your ads.
• Images – for Display ads, and even some search ads, you can change what images or designs appear.
• Video – these can appear as YouTube ads, you can test different lengths or video edits.
• Landing page – the actual URL on your website that users click through to.
• Bid strategy – whether you pay to aim for clicks, impressions, conversions or something else on your ads.
• Budget – how much you spend on your ads.
• Location – where your ads appear, by country, city, or even postcode.

There are even more aspects you could test on your ads. Importantly, when A/B testing, you should only test one aspect at a time, and keep the other parts of the ad the same, otherwise you won’t be able to determine which change caused an increase or decrease in your measurables. But let’s break down this initial list, to see how your ads could improve through A/B testing.

Should I A/B test headlines in my Google ads?

Headlines are the first thing potential users see when your google ad first appears. The language used here needs to entice a potential user to click on your advert.

They should relate to the search terms you expect people to use to find you, or relate to your product. But by A/B testing, you can determine which headlines users are responding more to.

Take for example, you have an advert for ticketing software. Test A will use headlines that describe the product you have directly such as:

• Software ticketing solution
• Ticketing software
• Software for tickets

This is direct, to the point and relates to the product. This no-nonsense approach could appeal to professionals who are looking for an efficient new system.

But, let’s say Test B uses headlines using more emotive and assertive language like:

• Need a new ticketing solution?
• The best software for ticketing
• The ticketing software you need

Would this appeal to someone looking for a new software system? You may have your own ideas about it, but until you A/B test you won’t know.

A/B testing can remove that guesswork, and sometimes end with surprising results!

Should I A/B test descriptions in my Google ads?

Descriptions add more context to your adverts; this is where you can explain your product or expand on a particular feature, promotion or accolades. When A/B testing descriptions you could start with A being the basic context of your product, such as:

• Reliable software you can trust, that opens in your browser.
• Subscribe monthly to our software or sign up for a free trial.

You could then have the variant, B, focus on reviews or accolades:

• “The most reliable software I’ve ever used” – John S.
• Winner of multiple awards in our field for customer service.

Once again, keep all other assets and budgets the same, and review which variant gave better results.

Should I A/B test keywords in my Google ads?

Keyword data is some of the most valuable to any business owner or marketer. By understanding what search terms your users are using to find and engage with your business, you can adapt your marketing to ensure you target the right people, for less budget.

Keyword testing can be tricky, as google allows you to input a large amount of keywords for your ads to appear for.

When keyword testing you can be very broad or super specific. Luckily, Google allows you to measure responses to individual keywords, which can make things easier to measure.

For the confident among you, you could test a single keyword or phrase per ad test, to see which terms lead to more clicks and conversions.

Should I A/B test images in my Google ads?

Testing image variants through your google ads is a very useful way of determining what your audience resonates with. Depending on your product, you could have a variety of images such as:

• An image of the product
• An image of people using the product
• People of a certain demographic using the product

By A/B testing images, you can understand your user behaviour in a more in-depth way.

Say, for example, you are testing out display ads for a particular product, and have images of the following:

• A man aged 25 – 30 using the product
• A woman aged 20 – 24 using the product
• An older couple using the product
• A family using the product

By A/B testing each of these images, you could gain valuable insight not just into what ads work better, but possibly even which demographic is more engaged with your product offering.

Images can be added to Search Ads as well, but do not always appear. But categorising images in a Performance Max Asset Group or Display Ads will allow you to see which of your imagery works the best.

Should I A/B test videos in my Google ads?

Video ads, which are usually YouTube ads, have a big impact on the success of your business. But there is a lot you could A/B test to determine which video works best for you.

If you are serious about video production and have the capacity to create multiple videos, then aspects you could A/B test include:

• Video length
• Order of sequences
• Voiceover accent or gender
• Typography
• Tone
• Calls-to-action

Do users respond more to a female voiceover than a male one? Do 15 seconds videos perform better than 25 second videos? Does a cheery, humorous tone engage more than a serious, corporate tone?

For most businesses, it is recommended to test video length of a YouTube ad at the very least, to see whether audiences engage more with shorter or longer ads. But if you have video editing capabilities to take it further, you could test until you have a video formula that works perfectly for you.

Should I A/B test landing pages on my google ads?

What if your ads are perfect, resulting in plenty of impressions and clicks, but you’re still not getting the right conversions?

It could be the page on your website that users are being sent to!

By creating a new landing page, with a different layout or call-to-action, you can run the exact same ad, to 2 different pages, and measure how users respond.

This won’t teach you more about what ads work best, but it will give valuable insights into what UX your customers respond more to once they get to your website.

Should I test bid strategy on my Google Ads?

This is getting into advanced territory, where you try to determine how best to use the budget of your Google Ads. Google offers various bid strategies, which are ways in which you determine how you will be charged for your ads, these include:

• Cost per click (CPC) – ask Google to stick to a budget of what you are willing to spend for each click on your advert
• Cost per mille (CPM) – ask Google to charge based on every 1000 views your advert receives
• Return on ad spend (ROAS) – ask Google to achieve a percentage of monetary sales based on how much you spend
• Cost per conversion – ask Google to spend £x for each conversion that takes place from your ad (lead form, purchase etc.)
• Maximise clicks – ask Google to focus on increasing clicks on your advert
• Maximise conversions – ask Google to focus on increasing Conversions on your advert

By testing different big strategies, you can determine how best to allocate your budget for a campaign. Does setting a £2 CPC rate result in better conversions than a £1 CPC rate? Is your ROAS target hit each time once it is set? Does maximising clicks result in more conversions than when you manually set a cost per conversion?

A/B testing your bid strategy can help you use your budget wisely, and save costs when done properly.

Should I A/B test budget on my google ads?

Some sectors are more competitive than others, some have higher ticket items that users spend more on. When this happens, the cost to successfully bid on keywords related to these sectors increases, and so a bigger budget is needed. Or is it?

By testing different budgets, you can determine whether you need to increase, or can afford to decrease, what you spend on Google ads.

The difference could be as simple as spending £50 in 1 month on one ad, and £100 in 1 month on another.

An increased budget could result in your outbidding competitors, reaching rank 1 on Google’s search results, and getting more clicks and conversion. Or you could find that the smaller budget results in a similar amount of clicks and conversions.

Testing the budget can help determine how much to invest in Google ads in future.

Should I A/B test location on my google ads?

Do you know where your customers live? Are they in a particular country, city, neighbourhood?

Location won’t be relevant to every business owner, but for those who are able to service a country, or even worldwide, testing location can determine which users are more likely to convert with you.

You can break this down by country, city, and even postcodes and streets. Perhaps you sell luxury items, and only want to target high net worth areas? Or you service a whole county, but don’t know which drive time towns are more likely to respond to you.

Testing locations won’t be necessary for everyone, but is just another example of how you can gain a more accurate insight into your audience on Google Ads.

Final Thoughts on A/B Testing

A/B testing is not a quick process. It can take months for every test to complete, but doing so is well worth the investment. By determining what ads and behaviours perform best, you can assign budget appropriately and get more results, for less overall costs.

If you are unsure on how to get started on A/B testing, or need help analysing the results you already have, then having a trusted digital marketing expert on hand is an absolute must.

Interested in finding out more about Google Ads and A/B testing? Talk to the team at Atomic Digital Marketing today.

A/B testing is what separates good campaigns from great ones. Our digital marketing services include paid media management built around continuous testing and optimisation — so your budget always works harder.