Changes in the Google landscape
The advertising world has changed a lot in the last couple of years. If you don’t have online presence, you hardly compete anymore. The impact of this is something we have been noticing on a daily basis with our clients as well. Due to the sharp increase in competition, it is becoming more and more difficult to stand out as a company. How do you avoid fishing in the same small pond as your competitors do? How do you distinguish yourself and start fishing in the great big ocean instead?
Google introduced the answer for this over a year and a half ago (February 2021): putting Broad match keyword types central. Moreover, the Broad Modifier matches were replaced with the existing Phrase match keywords. The behaviour of Exact match types wasn’t left untouched either. All of this with the aim to simplify campaign and keyword management and “help advertisers reach more relevant user searchers” source: Google Ads Developer Blog
A brief summary of the new behavior of the keyword match types:
- Phrase Match keywords are the new Broad Modifiers Matches (BMM); this new and improved Phrase Match behaves the same as BMMs, as it still observes word order when relevant to the meaning and, on top of that, also captures more volume than the original phrase match. Read Google’s article here.
- Exact keywords are no longer triggered based on highest ad rank, but the strongest matching keywords are now given priority in the matching process.
- With Broad keywords, additional signals (see image below) have been added for higher quality and more relevant matching to searches that are somewhat similar to your specified keyword: synonyms, related searches and other relevant variants.
Thus, with these changes to the different match types the focus shifted more and more to broad keywords. According to Google, it should be the solution to fish in a larger pond for more new and relevant customers. This type of keyword looks at user intent based on Artificial Intelligence, without the intent being specified in the search query. For example, in a simple search like “rain boots”, Google could distinguish between people with a buying intention and without a buying intention. However, it turns out that despite the changes to this type of keyword, broad matches in practice do not deliver on the promises of quality and relevance in many cases.
For example, in one of our client’s accounts, “gastric balloon” came up with the broad keyword “food slimming”. And “log cabin” was matched to the search term “decorative logs & stones”. As a result, much of the available budget was spent on irrelevant visitors, who were not looking for the products or services offered by the client. All of that despite the fact that Google aims to provide the most relevant content for the user doing a specific search and at a specific time.
In addition to the changes in match types, the market also changed significantly due to COVID-19:
- People have started searching differently as they are concerned about how inflation is impacting them — and how they can minimize the impact on their personal finances. For example
- Google data shows search interest for “why” in relation to “expensive” has grown 20% globally in the last 6 months alone
- Value-seeking behavior. Interest has also increased in 2022 for “cheap” searches.
- A 45% surge in search interest for the phrase “loyalty program”.
- Increase in searches for “second-hand” and “used”.
- Boom in reselling apps.
- Moreover there 15% more new searches a day.
- More parties have started selling online, or in other words competition is higher than ever and this means more who are fishing in the same pond.
So, despite decreases in relevance and quality, we still recommend to test with broad match keywords. Especially now that COVID-19 isn’t taking over our day to day lives that much anymore and the market demand/growth has stagnated, it is important to fish in a bigger pool of searches. Using only Exact & Phrase matches does not allow you to anticipate the changing market and might leave you missing out on many potential conversions. Broad Match gives the opportunity to move from a pond to the wide ocean allowing for much more potential to catch. In other words, Broad Match offers a greater opportunity to capture new relevant searches. But how do you start and how do you make sure you test it the right way, without affecting the performance of the campaigns too much?
Testing the effectiveness of Broad keywords
The answer is: test, test, test. But, before you start doing this, it is important that the campaign in which you are going to test it meets the following conditions:
- It uses an automatic / smart bidding strategy: Maximise Conversions, Target CPA, Maximise Conversion Value or Target ROAS.
- Enough data passes through the campaign.
- More than 3,000 impressions per week per ad group
- More than 15 to 30 conversions per month
- The campaign is not limited by budget
So how exactly do you test with Broad Match keywords? Google provides several options to take care of this with the simple click of a button.
- Auto-upgrade all keywords to Broad Match directly in the campaign via the Recommendations page.
- Select the best-converting keywords and convert them to Broad Match by changing the match type.
- Apply a One-Click experiment, which tests the original setup to conversions against converting all keywords to Broad Match.
However, we recommend to use a slightly different approach, as we want to maintain control over which keywords are converted to Broad Matches. This is because keeping some Exact keywords is important for relevance; broad keywords, after all, use other keywords in the ad group as reference material. And, as mentioned earlier, you don’t want to affect campaign performance too much. For this reason, we always first test the original against a variant in an experiment, or an A/B test.
Implementing Broad Match keywords in your targeting strategy?
How do you decide whether or not to add broad keywords to your campaigns? First of all, it is important to take a longer learning period into account. The system needs time to distinguish the difference between users who have a high probability of buying and users who don’t. Therefore we advise against analyzing and making adjustments too early on while the test is running, as the system will start learning over and over again. You can always stop the test in the middle if the results turn out to be very negative. However, we advise to leave the test running for a minimum of 4 weeks as only then will you be able to see the impact on for example Return On Ad Spend (ROAS) and costs. For example, for a large ecommerce client, the ROAS increased with 32% while seeing positive results on other important metrics as well (Image 4). For another (B2B) client, the amount of leads quadrupled. When finishing the broad match experiment, a search term report analysis is an important part in determining the success of Broad Match. Efforts should be done on excluding non-relevant keywords around competitors etc.
This is how you will succeed with broad keywords
- If you are reaching your goals with Exact and Phrase match keywords and you want to grow, you can test with Broad keywords. Make sure you are using a smart bidding strategy.
- A/B test with Broad keywords versus Exact & Phrase match before converting everything to Broad matches.
- Keep your Exact match keyword types in the campaigns, these will always take priority over Broad keywords and will always be more relevant.
- Start with a small budget to reduce costs while the algorithm goes through the learning period. If this proves successful after 4 weeks, you can increase budgets.
- Analyze the Search Term Report extensively, and make sure to add junk traffic as negatives to your campaign or in Negative keyword lists.
Do you also want to fish in a wider ocean and stay ahead of the competition, but need help testing and analysing Broad Match results? Or are your campaigns running on less budget, impressions and/or conversions, but still want to test with broad keywords? Then get in touch with our experts and we will be happy to help!