Google AI Mode Recipe Carousel Traffic: How to Check Yours

Google's March 4, 2026 AI Mode update links meal-idea searches, like "easy dinners for two," to recipe creator sites. A June 2026 update moved recipe links to the top of responses. Search Console has no AI Mode report, so segment meal-idea queries, compare clicks year over year, and write down what you find.

TL;DR:

• Google's AI Mode now surfaces recipe links for meal-idea searches, but it gives you no dedicated report to measure them.
• Search Console folds AI Mode into the Web search type, so you must segment by query type yourself.
• Meal-idea queries are the ones to filter; technique and substitution queries are a useful comparison group.
• Complete Recipe markup is a sensible baseline, though Google says no special schema is needed for AI features.
• Record your numbers today, because a measurement you did not save cannot be compared in November.

What did Google change in AI Mode recipe results in 2026?

Two announcements matter here, both from Google's Robby Stein.

On March 4, 2026, Stein said that for meal-idea searches like "easy dinners for two," AI Mode lets users tap a dish to see links to relevant recipe sites plus a short overview. Google also said it was adding more recipe details, including cook time, which testers found useful for deciding on a recipe.

On June 30, 2026, Stein announced a new visual treatment that puts recipe links at the top of AI Mode responses for relevant recipe queries. The cards show images, creator name, recipe ratings and number of ingredients.

That is the "carousel" most publishers are talking about. The word is shorthand. What matters for you is the behavior: a visual row of dishes, with links out to creator sites.

If you missed these, you are not alone. Recipe publishers have had a loud couple of years, and it is easy to tune out another Google announcement. This one is worth a look because it is about links to you.

Which recipe queries trigger the carousel, and which don't?

Google's announcement describes meal-idea queries. That is the only query type Google has named.

A vendor blog, Recipy, goes further. It reports that the update triggers on meal-idea queries and does not trigger on technique or substitution queries. Recipy is a vendor, not Google, and Google's own announcement only describes meal-idea queries. So treat the technique and substitution claim as a hypothesis to test, not a rule.

Here is how I would sort your queries to test it:

Query type	Example	Expected to trigger?
Meal idea	"easy dinners for two"	Yes, per Google
Technique	"how to brine a turkey"	Reportedly no (vendor claim)
Substitution	"buttermilk substitute"	Reportedly no (vendor claim)
Specific dish	"chicken tikka masala recipe"	Unknown, test it

The last row is the honest gray zone. Nobody has published a reliable rule for dish-name queries, so I would not guess. Put them in their own bucket and watch them separately.

Can Search Console show me AI Mode clicks?

Yes and no. The clicks are in there, but you cannot pull them out.

Google's documentation says sites appearing in AI features, including AI Overviews and AI Mode, are reported in the Search Console Performance report within the "Web" search type. There is no separate AI Mode view.

A few details help you read the data correctly. In Search Console, an AI Mode click counts when a user clicks an external link, and a follow-up question in AI Mode counts as a new query. That means one chat session can show up as several queries in your report.

GA4 has a similar blind spot. In GA4, Google's default channel grouping puts AI Overviews and AI Mode inside Organic Search. So neither tool will hand you a clean "AI Mode" number.

That is why segmentation is your best tool. You cannot see the feature, but you can see the query types it touches.

How do I segment Search Console by meal-idea queries?

Here is the method. It takes about 20 minutes the first time.

1. Open Search Console and go to the Performance report for Search results. Keep the search type on Web.
2. Add a Query filter. Use the custom regex option and enter a pattern that catches your meal-idea phrasing, such as dinner ideas|easy dinners|meal ideas|for two|weeknight|what to make. Adjust it to the way your readers search.
3. Set the date range to the period since March 4, 2026, and use the Compare tab to compare against the same dates in 2025.
4. Record clicks, impressions, CTR and average position.
5. Repeat with a second filter for technique and substitution queries, such as how to|substitute|instead of|can i use.

Step 5 is the part people skip. Your technique and substitution group acts as a rough control. If both groups move the same direction by a similar amount, the cause is probably broader than the carousel. If meal-idea queries diverge from the control, you have something worth investigating.

Two warnings. Your regex will miss queries, and it will catch a few it should not. Skim the query list for each filter and tidy it once. Also, do not change the pattern next month without noting it, or your comparison breaks.

What should I benchmark, and what can't the data prove?

Start with what the data cannot do. Because Search Console click totals cover all Web search activity, comparing page-level data cannot prove whether an AI impression did or did not send traffic. You can measure what moved. You cannot cleanly prove why.

That limit shapes what to benchmark. Here is what I would track:

• Clicks and impressions for meal-idea queries, year over year
• The same two metrics for your technique and substitution group
• CTR for both groups
• The date and the exact filter you used

Context helps too. Coverage of the broader AI era has been grim. After Thanksgiving 2025, Search Engine Land, summarizing Bloomberg, said many food creators reported 30% to 80% drops in Google traffic, with one blogger citing a 40% year-over-year decline. Those are self-reported figures, not audited data, so do not use them as a target. Bloomberg's reporting was based on interviews with 22 independent food creators, which is a small sample.

ALM Corp also cites a September 2025 Seer Interactive study finding that organic click-through rate on queries with AI Overviews fell 61%, from 1.76% to 0.61%. That is a secondary citation, so check the original Seer study before you quote it anywhere.

Notice what these numbers are. They are context, not benchmarks for your site. Your own year-over-year trend is the only baseline that matters.

Benchmark each recipe page against its own performance in the same weeks last year and against similar pages that AI features don’t surface, and focus on revenue and engagement per visit rather than raw traffic. Don’t over-read the new Search Console AI report, since it counts impressions rather than clicks and only has about five months of data, which overlaps a core update and seasonal swings. Sitewide averages, single-page spikes, and third-party “AI visibility” scores are also too noisy to trust on their own.

Which Recipe schema fields does the carousel appear to use?

The June cards show images, creator name, ratings and number of ingredients. Search Engine Journal noted that these details resemble fields typically supplied by recipe structured data, but Stein did not say whether structured data is mandatory for the AI Mode treatment.

Google is also clear on the other side. It says no special schema.org structured data is needed to appear in AI features.

So treat schema as hygiene, not a magic switch. Here is how the card details line up with Google's Recipe structured data documentation, which lists name and image as required and aggregateRating and recipeIngredient as recommended:

What the card shows	Likely related field	Status in Google's docs
Dish name	name	Required
Image	image	Required
Recipe ratings	aggregateRating	Recommended
Number of ingredients	recipeIngredient	Recommended
Cook time	A time field in your recipe card	Check your markup
Creator name	Your byline or author markup	Check your markup

The mapping in the first four rows follows Google's documentation and the card details. The last two are my inference, so check how your recipe plugin outputs them.

A quick audit: open three of your meal-idea posts in Google's Rich Results Test and confirm that name, image, ratings and ingredients all appear. If ratings are missing, that is the likeliest gap, since a card can only show a rating you have marked up and collected.

Which Thanksgiving posts should I check first?

I could not find sourced data on how Thanksgiving queries behave in the carousel, so I will not pretend to. Use your own data.

Start with your 2025 Thanksgiving winners. In Search Console, set the date range to November 2025 and sort your pages by clicks. Take the top 10 to 15. Then sort them into two piles:

• Meal-idea shaped: menu ideas, "easy Thanksgiving dinner for a small group," side dish roundups, make-ahead plans
• Technique or substitution shaped: brining, spatchcocking, gravy fixes, ingredient swaps

Check the meal-idea pile first. Those are the posts most likely to meet the carousel, based on the query type Google described. Review the schema, the rating count, the cook time and the lead image on each.

You will not know the full Thanksgiving effect until the season arrives. What you can do now is build a baseline from last November and fix the obvious gaps while there is time. If you want a planning frame for the weeks ahead, the seasonal content calendar walks through a 90-day plan, and refreshing your best older posts is covered in updating old recipe posts before writing new ones.

How do I build income that doesn't depend on one search feature?

The honest lesson of the last two years is that any single Google feature can change under you. A carousel that helps this quarter can shift next quarter, and you will find out from your analytics, not a warning email.

The sturdier move is to own part of your audience. An email list is the clearest way to do that, because it gives you a way to reach readers whether or not a search feature sends them. If you want a starting point, MailerLite is one option for email. MailerLite (affiliate link)'s strength is that it gives you a place to build and message a list you own. Its limitation is that it does not bring in subscribers by itself, so you still need a reason for readers to sign up, like a good lead magnet. The lead magnet ideas post is a good place to find one.

Wider diversification matters too. See how top food bloggers diversify traffic for the bigger picture. And if you are bracing for the ad calendar, Q4 ad revenue explains why the timing of traffic matters so much.

Do this today

Block out one hour and do this:

1. In Search Console, filter Web search queries for your meal-idea pattern.
2. Compare clicks from March 4, 2026 onward against the same dates in 2025.
3. Run the same comparison for your technique and substitution queries.
4. Save a screenshot and a one-line note in a spreadsheet: the date, your regex, clicks for both years, impressions, CTR and your read of the result.

That is the whole task. You are not trying to solve the question today. You are creating a record you can compare against in November.

FAQ

Does Search Console have a separate AI Mode report? No. Google's documentation says AI features, including AI Overviews and AI Mode, are reported in the Performance report under the Web search type. There is no separate AI Mode view, so you have to infer the effect by segmenting your queries.

Do I need special schema to appear in the AI Mode recipe carousel? Google says no special schema.org structured data is needed to appear in AI features. Search Engine Journal noted that the card details resemble typical recipe structured data fields, but Google has not said whether structured data is mandatory for this treatment. Complete Recipe markup is still a sensible baseline.

Can I prove the carousel caused my traffic change? Not from Search Console alone. Click totals cover all Web search activity, so page-level comparisons cannot prove whether an AI impression did or did not send a click. Treat your numbers as a directional signal and record them consistently.

Will technique and substitution posts show up in the carousel? A vendor blog, Recipy, reports that the update does not trigger on technique or substitution queries. Google's own announcement only describes meal-idea queries. Use that as a working hypothesis and test it against your own data.

Sources

• Search Engine Land, "Google AI Mode updates recipe results to better connect people with recipe creators" (2026)
• Search Engine Journal, "Google Moves Recipe Links To Top Of AI Mode Results" (2026)
• Google Search Central, "AI features and your website" (2025)
• Search Engine Journal, "Google Adds AI Mode Traffic To Search Console Reports" (2025)
• LLM Pulse, "Google Search Console Generative AI Report, Explained" (2026)
• Recipy, "Google's March 2026 AI Mode Recipe Update: The Real Playbook" (2026)
• Google Search Central, "Recipe (Recipe, HowTo, ItemList) structured data" (2025)
• Fortune/Bloomberg, "AI slop recipes are taking over the internet and Thanksgiving dinner" (2025)
• Search Engine Land, "Google and AI slop are ruining Thanksgiving for food bloggers" (2025)
• ALM Corp, "Google's AI Mode Recipe Update (March 2026)" (2025)

Recommended

• How to Diversify Food Blog Traffic: What Top Food Bloggers Do Differently in 2026
• A Seasonal Food Blog Content Calendar: How to Plan the Next 90 Days in One Sitting
• Stuck at the Same Pageviews? Update Old Recipe Posts Before You Write Another New One

Some links on CreatorPulse may be affiliate links. If you click one of these links and make a purchase, the company providing the product or service may pay CreatorPulse a referral commission.