When a responsive search ad headline does not show as you wrote it
A headline you put in the ad is missing from the search result. A responsive search ad shows a combination of headlines and descriptions, based on things such as how they relate to the query. A saved headline is not shown every time (Google Ads Help). That is the same mechanism whether the headline is a price, an audience, or a feature. It is also not a simple even random rotation.
This article uses our own ad, where a price headline was not selected. It is not an argument that the price itself was rejected. The step is to check which combination actually served, which query brought the person, and what the ad promised. After that, it is easier to decide whether to pin a headline or test another ad.
Separate the headlines you saved from the headlines that served
In the Google Ads ad screen, check the saved headlines, asset performance, and the combination report, in that order.
| What to look at | What to check |
|---|---|
| Saved headlines | Whether the roles are covered: kind of service, audience, benefit, price |
| Combinations | Whether any headline still makes sense alone and with the others |
| Search terms | Whether the ad that served matches what the person searched for |
| Landing page | Whether the page confirms what the ad promised |
In our search ad, the price headline was saved. The combinations we checked were abstract lines such as “help with improvement,” “decide the next move,” and “free,” and they did not say what the service was. Fixing only the price headline that was not selected does not fix a combination that is hard to understand.
Match the searcher’s purpose, not only the price
Review headlines by role: the problem they searched for, who the service is for, the benefit, and price or how easy it is to start. Google also recommends relevance to the query, a customer benefit, and concrete wording (responsive search ad best practices).
Then open the landing page and check what the stated price actually includes. For Shoryaku AI, paid plans start at ¥980 a month, and the 14-day free trial is the Standard plan. Detailed search-term and ad-asset analysis is on a higher plan. Avoid copy that can be read as “¥980 a month includes the detailed analysis” or “the free trial includes every feature.”
Two choices when one headline has to be said
One is to pin a headline you must say to a specific position. Pinning reduces the combinations that can serve. Google says wording that every ad must include can be pinned to headline 1 or 2, or description 1, and also says not to pin too much in the usual case (about pinning headlines). If you want to state a price, check which position you pin, and keep the description and the landing page aligned with it.
The other is to keep the current ad and add another ad, in the same ad group, that states the point more clearly, then compare them. We added an ad built around price and benefit, and pinned two price lines to headline 2. Each ad has its own traffic identifier, and the original ad is still there. This is not an experiment that splits delivery evenly.
Either way, do not treat the number of headlines, or the Ad Strength score, as the result. Google describes that rating as feedback while you create the ad, not as a measure that decides Ad Rank or the auction (about Ad Strength).
Do not call a winner from a few clicks
The day after we added the price ad, it had 55 impressions and 2 clicks. The original ad had 17 impressions and 0 clicks. The two clicks were for “ai マーケティング” and another company’s name. That does not show that a price message reached people looking for Google Ads improvements. Both ads had 0 conversions, and we had not confirmed that the signup-complete tag was received on the Google Ads side.
We also changed the ad text and the ad group later. Impressions and clicks from before and after that change cannot be pooled to say the price ad won. When you compare, split at the change date, and look at the search terms, impressions and clicks by ad, signup starts, and signup completions in the same period. If measurement is unconfirmed, check that the tag is received at the completion point first.
Check the ad copy in ChatGPT or Gemini
Before you rewrite the copy, paste this. If a field has no search terms or results, write “unknown.” Do not let it fill in numbers.
Review this responsive search ad.
Product, audience, and the price and plan scope:
Main search terms in the ad group:
Saved headlines and descriptions:
Combinations that actually served:
Impressions, clicks, and completed signups by ad (same period):
Landing-page headline, price, and free-trial explanation:
Whether the signup-complete tag was received:
1. From the combination alone, can you tell whose problem, and which problem, the service addresses?
2. Do the price and the free trial match what the landing page actually includes?
3. Which headline has to be said every time? If you pin it, also state the cost of fewer combinations.
4. If you keep the original ad and test another, limit the change to one hypothesis.
5. If the data is thin, withhold a winner and name the next metric to watch.
Do not invent results or advantages that are not in the input.
In Shoryaku AI you can line up the Google Ads headlines, facts from the landing page, and the search terms, and look at an improvement. Ad settings do not change on their own. A matching change is applied one at a time, after you confirm it. Start on the Google Ads analysis page and check that the ad and the page promise the same thing.