"While listing ads get clicks, they do not lead to conversions (CV)" and "the CPA (cost per acquisition) continues to rise." — Many advertising operators face this issue. What is the root cause of this problem?

In many cases, the core of the problem lies in the "keyword selection" strategy for listing ads. Simply setting keywords with high search volume tends to increase irrelevant clicks and waste valuable advertising budgets.

This article thoroughly explains seven essential tips for keyword selection to maximize the effectiveness of listing ads. By grasping these points, you will be able to optimize your advertising budget and build a keyword strategy that connects you with true customers.

1. "Preparation is 80%": Strategic design to finalize before relying on tools

Effective keyword selection is determined 80% before opening keyword tools. Even professional operators often overlook this preparation, leading to a waste of advertising costs. Successful campaigns start with strategic design, not tools.

Diving Deep into the Three Elements: "Product, Persona, Search Intent"

Before relying on tools, it is essential to thoroughly analyze the following three elements.

• Understand the product from multiple angles: List all the "functional values (low price, high performance, etc.)" and "emotional values (sense of security, time-saving, etc.)" of your own products or services. This allows you to hypothesize what challenges customers are trying to solve and what words they might use to search.

• Specify the persona: Clearly define the ideal customer image you want to deliver your ads to (age, occupation, challenges they face, motivations for seeking services, etc.). Clarifying the persona significantly improves the accuracy of keyword granularity and match type selection.

• Dive into search intent (Why): For instance, behind the search phrase "SEO tool comparison," there are various backgrounds, such as wanting to understand the differences between paid and free tools or checking functional disparities. Formulating hypotheses about why users searched with that keyword, including their purpose and emotions, becomes the foundation for accurate keyword selection.

In keyword design for listing ads that lead to results, it is paramount to apply "strategic thinking" before searching with tools.

2. Break free from dependence on big words and improve cost-effectiveness

Many operators fall into the trap of concentrating their budget on "big words" with high search volume, like "insurance" or "English conversation." These seemingly attractive keywords carry two significant risks.

1. High competitiveness and CPC: Competition is fierce as many companies advertise, causing click costs (CPC) to rise.

2. Ambiguous search intent: The searchers' intent is too broad, leading to a large number of clicks from users, such as students gathering information, who do not lead to conversions, ultimately driving up the CPA.

The solution is to shift focus to more specific and high-intent “middle words” or “long-tail words.”

• Big words (not recommended): English conversation

• Long-tail words (recommended): English conversation online free trial for working adults

By narrowing down to keywords with clear search intent, you can optimize your advertising budget, achieve a high conversion rate (CVR), and establish the basic principle for improving cost-effectiveness.

3. The key to CPA improvement lies in eliminating "wastage" with excluded keywords

In CPA improvement, the most impactful yet often overlooked aspect is the setting of "excluded keywords." This setting should be viewed as a crucial process to be considered in two stages: initial "preemptive defense" and ongoing "continuous improvement" after starting operations. Neglecting this leads to ads being displayed for search phrases unrelated to your services, resulting in continuous wastage of click costs.

Below are examples of keywords commonly excluded as preemptive defense.

• Words for information collection purposes: free, how to use, meaning, what is

• Words with low relevance: used, job postings, comparison

• Groups to avoid with BtoB products: part-time jobs, students

A slight difference in search intent can lead to significant losses. For instance, when a real estate company displays ads for "buying used apartments" (for companies looking to buy properties), if the ads show up for general users searching for "buying used apartments" (who want to purchase properties), it wastes advertising costs on completely opposite needs. To prevent such wastage, accurately excluding keywords with opposing intents to your services is essential.

Setting excluded keywords is not a one-time task. Regularly checking the "search phrase report" and continuously excluding phrases that do not lead to results through diligent maintenance is the shortest route to CPA improvement.

4. In the AI era, account structure relies on "data volume"

In today’s world of automated bidding through AI, the purpose of account structure is no longer just to organize for human manageability. Its primary goal is to become a disciplined data provider for AI.

To maximize the learning effect of AI, it is recommended to secure more than 30 pieces of conversion data per campaign per month.

If the conversion data does not meet this standard, AI will not be able to learn adequately, and optimization precision will decline. In such cases, the following measures are necessary.

• Setting micro-conversions: Instead of only setting the final conversion (such as purchase completion), set intermediate goals like accessing a form or downloading materials as "micro-conversions." This increases the signals (data) for AI to learn. However, care must be taken to monitor whether AI is overly optimizing towards these intermediate goals, as there is a risk of gathering only users with low purchase intentions, potentially degrading user quality.

• Campaign integration: While it is important to subdivide keyword intents into finer ad groups, if the data is too dispersed, the efficiency of AI’s learning will decrease. If conversion numbers are low, a judgment call may be needed to integrate multiple campaigns or ad groups to aggregate data, thereby accelerating AI learning.

Sufficient conversion data (more than 30 per month) is necessary for automated bidding to function effectively.

5. Quality Score is the strongest weapon to overturn "bid amount"

When improving the efficiency of listing ads (CPA and ROAS) from within, the most crucial metric is the "Quality Score." The ad position is not determined solely by the height of the bid amount.

The ad rank is generally determined by the following formula.

Ad Rank = Bid Price × Quality Score

This mechanism allows ads with a high Quality Score to appear in higher positions with lower bid amounts compared to competitors.

• Company A: Bid amount 300 yen × Quality Score 5 = Ad Rank 1,500

• Company B: Bid amount 200 yen × Quality Score 8 = Ad Rank 1,600

In this example, Company B surpasses Company A in ad rank despite having a 100 yen lower bid amount due to its higher Quality Score. In other words, the Quality Score is the strongest weapon to keep advertising costs low.

Three factors that make up the Quality Score

To improve the Quality Score, it is necessary to address the following three elements.

• Estimated Click-Through Rate (eCTR): The likelihood of an ad being clicked when displayed.

  • Measures: Create ad text that includes clear user benefits, such as "free consultation" or "next-day delivery," which induce users to want to click.

• Relevance of the ad: The degree to which keywords, ad text, and user search intent match each other.

  • Measures: Group keywords with similar search intent into one ad group and ensure that the group’s main keywords are always included in the ad headlines.

• LP usability: Whether the landing page that users reach after clicking the ad is beneficial and easy to use for them.

  • Measures: Design the LP so that the information promised in the ad can be found immediately, maintain page load speed within three seconds, and ensure full compatibility for mobile viewing.

6. Keyword selection is not a one-time task

Keyword selection is not a static task done only once when starting the ad. As market demands and user search behaviors constantly change, continuous improvement (PDCA cycle) is the key to success. This is the "continuous improvement" phase in the excluded keyword strategy discussed in section three.

The primary improvement activities to undertake after starting operations are as follows.

• Regular performance analysis: Continuously monitor key indicators such as CTR (Click-Through Rate), CVR (Conversion Rate), CPA, and Quality Score, and stop or reassess keywords that are performing poorly.

• Utilization of search phrase reports: This report is a treasure trove that shows the actual phrases users searched for. Primarily, two actions are executed.

  1. Exclusion of non-performing phrases: Quickly identify search phrases that are being clicked but do not lead to any conversions and add them to the excluded keyword list.

  2. Addition of high-performing phrases: If you discover search phrases that unexpectedly lead to high conversions, officially add them as new keywords. This allows for optimizing bid prices and ad text for that keyword.

In an era where automated optimization by AI is advancing, the ability of operators to manually refine keyword sets based on data is what separates successful campaigns from unsuccessful ones.

7. Responsive Search Ads (RSA) convey intent through "fixed features"

Responsive Search Ads (RSA) are a powerful format that combines multiple registered headlines and descriptions automatically by AI to deliver optimal ads. This offers the advantage of accommodating diverse user needs.

However, this automation also carries the significant risk of "unintended ad text being displayed." For example, there is a possibility that combinations contradictory to the brand message or ads with awkward contexts may appear.

The function to control this risk is the "fixed setting function (pinning)." Google recommends giving AI maximum freedom, but there are times when it is necessary to prioritize brand safety and message clarity at the expense of algorithmic optimization.

Using this function, specific headlines or descriptions can be displayed in specific positions (such as headline 1, headline 2, etc.). By ensuring that critical messages, such as brand names, legally required disclaimers, and the most critical strengths, are fixed, you can secure message consistency and brand safety. Operators are required to regularly check the "combination" reports to ensure that no unnatural ad text is generated, while flexibly revising their fixed strategies.

Conclusion

To achieve results with listing ads, it is essential to collaborate with AI rather than blindly trusting it. Success comes from the synergy between human strategy and machine execution capabilities. The role of the operator is to perfect high-quality "inputs" so that AI can perform at its best. Through deep persona analysis, correct account structure, and high-quality advertising assets, you can guide AI correctly and extract the optimal "outputs."

The goal should be to move away from simply "buying clicks" and instead "investing in targeted dialogues with potential customers".

If you are looking to implement the principles outlined in this article in a more sophisticated and efficient manner, the AI marketing optimization platform "Cascade" automates continuous improvement cycles, such as analyzing search phrase reports and excluding low-performing keywords, allowing you to focus on more strategic tasks. If you are interested, please take a look at the details.