At the beginning of this year, Google announced what to expect from Google Ads in 2022 and how to deal with the changes that will be in place throughout the next 12 months. One of the main topics is the recommendation of using broad match keywords and how they can work well when combined with responsive search ads (RSA) and a smart bidding strategy.

Of course, most PPC practitioners feel shivers down their spine every time they hear Google recommending using broad match keywords and, even though Google is arguing that improvements were made on the keyword matching technology that triggers the brand’s ads for relevant queries, there is no PPC manager that likes the idea of letting Google have control of which queries the ads will be displayed to.

On the other hand, more and more, marketers are having to scratch their heads when faced with the challenge of increasing impression share for their Google search campaigns without compromising the relevance of the audience triggered by the search queries.

But why is Google giving so much importance for the broad match keywords now? And why do they claim that the combination of broad match + responsive search ads + automated bidding can deliver a more successful ROAS performance?

In this article, we’ll be going to explore each element to give you an idea of the pros and cons of applying Google’s recommendation.

Golden Triangle a image

Why broad match type keywords can become the real deal

The challenge when creating campaigns on Google Ads is finding the right balance between not being too broad, neither too narrow. And here is where it starts to become tricky about choosing your KWs match type: if you opt to use only phrase and exact match types, you may end up with an audience pool that is too small; whereas adding broad match keywords can trigger your ads to loads of irrelevant search terms thus wasting money and impacting your CTR.

But let’s say your aim is to use smart bidding strategies to decide the ideal bids for your search KWs towards your main KPI. Would you rather give the machine learning a refined and niche pot of search terms to optimise, or a larger range of search terms where the machine learning can find the most cost-efficient search ones, regardless of their stage in the keyword funnel?

Whilst the first option sounds safer, the second option will give you amplitude and opportunity to scale your ads and reach users beyond the keywords added in the ad group. This is because by using broad match keywords, you’re telling the machine learning to optimise the performance by focusing on the queries instead of specific keywords and, when combined with smart bidding strategies, will allow the algorithm to identify the search queries that can lead to your goals and optimise your bids towards it.

And what about the campaign structure?

If the idea is to use broad match keywords and trigger search ads towards queries, there’s no need to have campaigns (or ad groups) for each match type ad groups (broad, phrase and exact). This is a major change on how PPC practitioners currently breakdown their campaigns into ad groups.

It doesn’t mean this a bad thing though. A more simplified campaign structure will allow PPC specialists to focus the split of their campaign ad groups on themes and can save significant set up time from complex campaigns.

Below is a graph of what Google is recommending being an optimal campaign structure, when using broad match + RSA + smart bidding strategy:

Golden Triangle b image

It’s noteworthy that Google is recommending keeping a DSA ad group in the same campaign with standard ad groups, which can sound a little bit odd as you may think that the DSA ad group will end up cannibalising the campaign media spending, not allowing other ad groups to shine.

However, having a DSA ad group together with other ad groups in a campaign that uses smart bidding strategy will give the machine learning more data to optimise the ads delivery towards your campaign goal.

Why are responsive search ads so important?

According to Google, from July 2022 onwards, responsive search ads (RSA) will be the only option to create ads in Google search campaigns. It’s also known that RSA helps advertisers to create flexible ads that adapt to device widths and reach more potential customers with multiple headline and description options giving the search ads the opportunity to compete in more auctions and match more queries.

Therefore, if you’re considering creating a search campaign using broad match keywords to expand your campaigns’ reach, alongside with smart bidding strategy to have the machine learning setting the right bids to your keywords, why not add a search ad with multiple assets options and allow Google to determine the best ad copy combination for each query?

Of course, using responsive search ads can seem limited regarding ad copy testing, however having a good RSA can be enough to help increase the relevance of your ad copy towards the search query… and we all know that better quality score reduces the cost per click.

So having RSA combined with smart bidding campaigns will not only prevent you from spending time setting up multiple ad variations and extensively analysing performance, but it will also give Google’s machine learning the ability to optimise ad copy combinations to get the best-performing assets.

How many conversions do you need to successfully use smart bidding?

If you have ever heard about Google smart bidding before, you probably also heard that campaigns start off in a learning phase, and until a minimum number of conversions are reached, the campaign will remain there until the algorithm has enough data to identify patterns and optimise.

When choosing a campaign that you would like to test the waters on using broad match + RSA + an automated bidding strategy make sure that you (or your PPC agency) are selecting a campaign with enough conversions in the past 30 days, or the campaign won’t have enough data to be effectively and successfully optimised.

“The recommend a minimum of 15 conversions in last 30 days to use their Golden Triangle (broad + RSA only + smart bidding) BUT 50 - 100 is actually better IF you want this method to outperform manual strategies.”

NeptuneMoon

Quote Author: via Official PPC Chat

Is Google migrating to an ‘automated’ environment?

All the updates seen in Google Ads over the past year indicate that Google’s machine learning is improving and is more capable of taking the right decision towards which queries to show your ads against and what the ideal CPC bid should be to achieve your goal.

Also, Google is not the first digital advertising platform to promote automation instead of manual control. Facebook, LinkedIn and Amazon Ads are ahead of the game in this matter, and these tools have been proving the value of having efficient machine learning to optimise the key elements of your online campaigns (like search queries and bids), allowing digital marketers to dedicate their time with strategic decision making.

Furthermore, Google machine learning is capable of dealing with thousands of signals in real-time, with these signals going beyond queries and bids, providing better target accuracy and more predictive signals. This is not only intended to favour advertisers but also the user experience in the SERPs. Therefore, it’s believed that Google will prefer to have its machine learning in place instead of manual controls, aiming to create a more balanced and better environment for advertisers and users.

What should digital marketers and PPC practitioners do then?

You may have now gotten super excited with the fact that the combination of broad match KWs + RSA + smart bidding strategy can help display your ad in front of as many eyeballs as possible, with very minimal effort, and use large amounts of data to optimise and buy the most efficient traffic.

However, this doesn’t mean that you need to go to all your Google search campaigns and start adding or changing existing KWs match types to broad match, and/or start setting smart bidding strategies across the board.

It's recommended that you run an experiment campaign at first and measure the impact in the performance of key metrics, like CTR, CPC, conversion rate and CPA. It’s reasonable that you give enough time to your experiment campaign to gather substantial performance data before taking any major decision.

Another important thing is to make your colleagues or clients aware of what you’re going to do and set the right expectations. It’s very likely that with broad match KWs in place, you will see an increase in the paid traffic going to the website (which may or may not convert well) and nobody wants anyone getting too excited or too scared with the change.

Once you are confident on the data gathered from the experiment campaign, you can consider applying (or not) the changes into the original campaign.

Conclusion

You’re likely to be either confident and ready to rock with Google’s ‘Golden Triangle’ or panicking and trying to figure out where to start.

If you’re the latter, you need to be aware that some of these changes are inevitable (like the roll-out of RSAs), but there’s no reason to panic. The best thing to do is test the approach starting with one campaign at a time.

There’s also no need to rush and dive straight in. The earlier this approach is tested, the sooner you’ll be able to evaluate performance and potentially get ahead of the competition, but it’s important to understand whether the shift in approach is cost efficient or not. We believe for some advertisers it won’t be, or at the very least it won’t become the dominant approach.

If you’re not comfortable with using broad match keywords or may think that managing this recommended structure can be time consuming and/or ‘dangerous’, get in touch and we’d be happy to have a conversation about your account.

We’ll be testing this approach with our ad campaigns over time and will share any useful insights in due course on our blog or within our case studies.

Back to blog
Meet the author ...

Paulo Santoni

PPC Manager

Paulo has worked in PPC management for a decade and counting! From Google Ads to Programmatic media, from expectations management to performance delivery, from data analysis to ...