When creating campaigns in Google Ads and Yandex.Direct, it's necessary not only to choose the relevant keywords but also to group them correctly. Creating unique ads for each keyword can be time-consuming. Additionally, it will be difficult to track and analyze them since you'll have to monitor each one separately. Therefore, it's better to segment the keywords to streamline your work and enhance the overall campaign effectiveness.
Why is the "one keyword - one ad group" strategy not always effective? On one hand, this approach can yield a higher CTR, as unique ad texts and headlines are created for each keyword. On the other hand, these keywords are often low-frequency.
As a result, in Yandex.Direct, they receive a "Low impressions" status, and in Google Ads, when there are no impressions, such keywords are recommended to be deleted. Even adding a broad high-frequency keyword to the ad group might not rectify the situation: we will end up with a higher bid in the auction compared to the desired keyword. Moreover, if you have multiple ad groups with low impressions in the same campaign, adding the same broad keyword will lead to competing against oneself.
As a result, in Yandex.Direct, they receive a "Low impressions" status, and in Google Ads, when there are no impressions, such keywords are recommended to be deleted. Even adding a broad high-frequency keyword to the ad group might not rectify the situation: we will end up with a higher bid in the auction compared to the desired keyword. Moreover, if you have multiple ad groups with low impressions in the same campaign, adding the same broad keyword will lead to competing against oneself.
Why should keywords be grouped?
When you group keywords for ads, you gain several advantages:
- Enhanced Relevance: Grouping keywords allows you to tailor their display by phrases. If a user inputs one of the grouped keywords, they might see all ads from that group.
- Simplified A/B Testing: Create keyword groups, distribute them to different landing pages, and determine which queries yield the most effective results.
- Refinement of Search Queries before Campaign Launch: By categorizing keywords into groups (by theme or query volume), you can pre-select the most suitable ones for a specific advertising campaign before it even starts. This leads to cost savings on testing.
- Facilitated Analytics: Analyzing keyword groups is more convenient than examining each individually. You can identify ads with the highest CTR, pinpoint successful queries, and scale your efforts accordingly.
- Campaign Management Convenience: Grouping keywords reduces the number of ad groups, decreasing the likelihood of missing out on unprofitable or profitable campaigns.
- Prevention of Cannibalization: This is advantageous for those promoting different pages of a single website through context. By forming keyword groups for ads, you eliminate the scenario where multiple ads lead to the same page, ultimately competing against each other (which you're paying for).
Principles of grouping
Keyword clustering occurs based on two main principles:
Semantic Relevance: In this case, keywords are grouped together based on their semantic similarity. In such a group, there's a main word that reflects the common attribute or characteristic of the group. Other related keywords are grouped under this main word. For instance, if you're promoting an offer from a bookmaker company, you can cluster keywords semantically: sports betting, betting on sports, bets in bookmakers, and so on.
Shared Attribute: Depending on the collected keywords, you can segment them based on a specific attribute, such as geographic location or user intent expressed in their queries. Continuing with the betting offer example, depending on the geographical target, you can group keywords like: bets in Moscow, boxing bets in Moscow, where to place bets in Moscow, and so on.
Furthermore, after gathering keywords and distributing them into segments, you can perform additional clustering based on syntax. For instance, combining words like "buy" and "Buy" into one group. This allows your ads to be shown to users who type queries in different case formats. However, it's recommended to do this if you have a substantial number of keywords; otherwise, it could lead to too many groups, complicating the analysis and data collection for advertising campaigns.
Semantic Relevance: In this case, keywords are grouped together based on their semantic similarity. In such a group, there's a main word that reflects the common attribute or characteristic of the group. Other related keywords are grouped under this main word. For instance, if you're promoting an offer from a bookmaker company, you can cluster keywords semantically: sports betting, betting on sports, bets in bookmakers, and so on.
Shared Attribute: Depending on the collected keywords, you can segment them based on a specific attribute, such as geographic location or user intent expressed in their queries. Continuing with the betting offer example, depending on the geographical target, you can group keywords like: bets in Moscow, boxing bets in Moscow, where to place bets in Moscow, and so on.
Furthermore, after gathering keywords and distributing them into segments, you can perform additional clustering based on syntax. For instance, combining words like "buy" and "Buy" into one group. This allows your ads to be shown to users who type queries in different case formats. However, it's recommended to do this if you have a substantial number of keywords; otherwise, it could lead to too many groups, complicating the analysis and data collection for advertising campaigns.
Grouping manually
Collecting and segmenting keywords manually is worth considering when dealing with a small number of queries.
This approach requires time. However, humans always have a better understanding of word meanings compared to programs or algorithms, allowing you to more accurately determine keyword relationships. For ease of work, you can load all gathered queries into an Excel spreadsheet and set up a text filter based on the presence of specific words.
This approach requires time. However, humans always have a better understanding of word meanings compared to programs or algorithms, allowing you to more accurately determine keyword relationships. For ease of work, you can load all gathered queries into an Excel spreadsheet and set up a text filter based on the presence of specific words.
Attempting to automate
When dealing with a large number of keywords, it's advisable to utilize various software and services. They will save you time, but remember that they do not provide a 100% guarantee of accurate data processing. Therefore, always review and adjust the results. For automated keyword grouping, you can make use of tools like Key Collector, Rush Analytics, PPC-Help, CloudLemma, and more.
What should you keep in mind?
- Group keywords in sets of 3 to 5.
- Include at least 1 high-frequency keyword in a group with low-frequency keywords.
- Ensure that keywords do not appear in multiple groups simultaneously.
- Incorporate at least 1 keyword in the ad headline.
- Set up dynamic keyword insertion in the headline for rotation.
Useful tools
Just-Magic. The service enables keyword collection and segmentation according to Yandex and Google top searches. Additionally, it can identify query themes and suggest related phrases and words. While there is no free access, you can pay for individual tasks or opt for a monthly subscription.
SEMparser. A paid online service for automatic keyword grouping. After registration, there's a trial access for testing the service with 50 queries. Subsequent usage is billed separately, with the cost depending on the quantity (0.6 rubles per query for purchases up to 500, and so on).
Serpstat. A platform offering a variety of tools. It allows keyword research and segmentation for both Yandex and Google. The service is paid, but there's a demo version available to explore its capabilities.
Coolakov.ru. A set of useful tools with free access. The access is free, but if you need to group more than 500 keywords, you'll need to pay 0.2 rubles for each one.
SEOquick. A service for SEO optimization and contextual advertising with free access. It enables keyword collection, grouping, and conversion of keywords to Google Ads match types. The clustering limit is 40,000 keywords.
SEMparser. A paid online service for automatic keyword grouping. After registration, there's a trial access for testing the service with 50 queries. Subsequent usage is billed separately, with the cost depending on the quantity (0.6 rubles per query for purchases up to 500, and so on).
Serpstat. A platform offering a variety of tools. It allows keyword research and segmentation for both Yandex and Google. The service is paid, but there's a demo version available to explore its capabilities.
Coolakov.ru. A set of useful tools with free access. The access is free, but if you need to group more than 500 keywords, you'll need to pay 0.2 rubles for each one.
SEOquick. A service for SEO optimization and contextual advertising with free access. It enables keyword collection, grouping, and conversion of keywords to Google Ads match types. The clustering limit is 40,000 keywords.
Tips from Vlad Marlon of Marlerino Group
By what criteria to group the keywords?
It all depends on the collected semantics, trends, and keyword competition within the niche. There is no universal formula. When we gather keywords, we usually group them by categories and frequency, striving to maintain a balance between logic and numbers.
How many keywords can be in a group?
Again, it depends on semantics and how narrowly you want your ads to be displayed. The most ideal scenario is 1 keyword (including brackets and quotes) → 1 group → 1 ad. This significantly increases the workload, but based on our experience, it's still the most appropriate approach. If desired, you can launch up to 100 keywords in one group, particularly to validate their effectiveness. However, this usually comes with an internal sense of algorithm performance.
Tips and recommendations
One piece of advice — don't give up. It all depends on the day (lunar cycles) and the mysteries of advertising algorithms. Today, your ads might not perform well and yield no results, but tomorrow, without making any changes, you might achieve excellent results. It's crucial to establish your own working system and stick to it, consistently striving to improve based on the impact of actions on outcomes.
Conclusion
Keyword grouping is not as challenging a task as it might seem at first glance. However, the right approach to it can enhance the relevance of your ads to user queries and consequently increase conversion rates. Therefore, it's essential to prioritize this process when preparing advertising campaigns in Yandex Direct and Google Ads or optimizing your website for search engine promotion.