Local AI Overviews: how to get your e-commerce product cards cited (the Local Pack is no longer enough)
Summarize this article with AI
A Tuesday morning of panic: local traffic collapses
A client calls me on a Tuesday morning.
He has 47 retail locations in France.
Each store has its page on the e-commerce site.
14 months ago, these pages generated 4 200 organic sessions per month.
Today, 1 100.
−74 %.
Not a gradual decline. A sharp drop in March 2026.
The Local Pack ranked him first on « hardware store [city]. »
The clicks, though, weren’t following.
I look at a SERP live with him.
Above the Pack, an AI Overview block.
It answers what the searcher was looking for.
Without clicking.
Without calling.
His ranking wasn’t the issue. The answer surface Google was now displaying was.
The numbers confirm the alert. A Whitespark study, reported by Search Engine Journal, shows that AI Overviews appear on 68 % of local queries.
The Local Pack only shows up in 39 % of cases.
A 29-point gap.
And that gap is widening.
For my client, existing local pages had become invisible.
They generated authority.
They fed the Pack.
But the AI wasn’t citing them.
And without citation, no more traffic.
We stopped focusing on the Pack.
We attacked the problem at the source: how to make the AI Overview mention his cards.
Les chiffres de Whitespark sont sans appel : l’AI Overview s’impose dans plus de deux tiers des requêtes locales, tandis que le Local Pack seul recule. Visualisons cet écart pour comprendre l’urgence d’adapter vos pages.
AI Overviews vs Local Pack : quel format domine les recherches locales ?
68 % des requêtes locales affichent un AI Overview, contre 39 % uniquement le Local Pack
68% vs 39%: numbers that shake up local search
AI Overviews are no longer a beta experiment.
According to Whitespark’s study, analyzed by Adam Heitzman in Search Engine Journal, they dominate 15 markets.
The proportion varies by intent.
Simple transactional queries, like « tacos San Francisco, » still often trigger the Local Pack.
But as soon as the question carries hybrid intent, the AI Overview takes over.
Example: « average cost of a dental implant in Phoenix. »
AI Overview present in 97 % of cases.
With a mini Pack sometimes embedded.
Another example: « how long does an eye exam take near me. »
Display rate: 92 %.
Google synthesizes the answer.
The sources used appear as links.
If your product page or local card isn’t one of those sources, you lose everything.
The same pattern touches e-commerce.
A buyer types « best firm mattress available today in Lyon. »
The AI summarizes features, availability, reviews.
Three well-structured product cards feed the synthesis.
If you’re not there, the customer clicks on a competitor.
Or they order directly via the mentioned card.
Remember this: local visibility is no longer measured solely by Pack position.
You need to be the site the AI uses to answer.
Why your local pages are invisible in the AI Overview (even though you rank first in the Pack)
The local Pack relies on three signals: proximity, prominence, relevance.
The AI Overview works differently.
It chooses content that answers quickly and with authority.
Whitespark showed that cited sources combine:
- direct answers to the implicit questions in the query
- semantic structure that language models read well
- thematic authority established via relevant internal links
I use the DOSE framework I learned with Guillaume Attias at BMO Academy.
It builds this thematic authority.
Detect, Organize, Structure, Elevate.
Locally, it goes like this:
Detect hybrid intents around your products and store locations.
Organize these intents into semantic clusters.
Structure each local page to answer the cluster precisely, with marked-up data (FAQ, HowTo, LocalBusiness).
Elevate the whole thing by linking each card to a pillar product page or buying guide.
Your local pages become the terminal link of a semantic silo.
Not isolated islands.
AI Overviews choose sites that contextualize the answer within a network of reliable content.
If your « Paris 11 store » card isn’t linked to anything, the AI ignores it.
If it’s the landing point of an architecture that starts from a category page, goes through a comparison, then lands on local availability, it becomes a natural citation link.
Pour remonter la pente, nous avons suivi le framework DOSE de BMO Academy. Voici les étapes clés qui ont permis de passer de –74 % de trafic à +340 % de mentions dans l’AI Overview.
Le framework DOSE : 4 étapes pour structurer vos pages locales
Détecter, Organiser, Structurer, Élever – le processus appliqué au cas client
Concrete case: 47 store locations, +340% AI mentions in 3 months
Back to my client from Tuesday morning.
47 local cards.
Each card detailed the store: address, hours, directions map, 4–5 featured products.
Standard.
But all these pages were built like brochures.
No structured answers to questions customers ask.
No data exploitable by the AI.
First step: we audited the 47 pages.
We identified 12 recurring hybrid intents.
Like « parquet installation cost in [city]. »
« pergola delivery time to [city]. »
We built a silo.
A pillar page « Parquet Installation Guide » was written.
It answered generic questions.
Each local card received an FAQ block with local variations.
We marked up the FAQ with FAQ schema.
We added enriched LocalBusiness schema (price, trading area, services offered).
We connected the 47 cards to the pillar via star-pattern linking.
In 3 months, the results:
On 12 tracked hybrid queries, the client’s pages went from 4 citations in AI Overviews to 14 citations.
That’s +340 %.
Direct traffic from clicks on links in the AI Overview went from 120 visits/month to 390.
Modest in volume.
But highly qualified.
More importantly, classic organic traffic rebounded by +22 %.
Why?
Because thematic authority and freshness signals fed traditional SEO.
The AI Overview acted as an amplifier.
The restructuring transformed isolated pages into reliable sources for artificial intelligence.
Action plan to get your e-commerce cards cited in the AI Overview
The steps are concrete.
They don’t require extra ad budget.
They require structural discipline.
1. Audit hybrid intents in your local market.
Use a tool like Semrush or Ahrefs to extract queries « cost + [product] + [city], » « delivery time + [product] + [city], » « best + [product] + [city]. »
2. Build a semantic silo with a pillar product page.
Write a complete guide page on the subject.
It becomes the source of truth for the AI.
Each local card becomes a variation of this pillar.
3. Mark up each local card with structured data.
At minimum: LocalBusiness, FAQ, and Product if you list references.
Add the « areaServed » property to clarify your zone.
4. Integrate short, precise answers in the page body.
The AI Overview summarizes the first paragraphs that answer directly.
For example: « The delivery time for a pergola in Lyon is 48 h, subject to stock availability at your Lyon Part-Dieu store. »
5. Align your Google Business Profile data.
Hours, services, and reviews must match your website.
The AI cross-checks these sources.
6. Track AI Overview mentions.
Use SE Ranking or custom scripts to monitor your pages appearing as a source in the AI Overview, query by query.
You’ll know which pages are being picked up.
The trap to avoid: keep producing without restructuring
Many e-commerce businesses add pages.
Pages by city.
Pages by category.
Duplicated, thin, unlinked.
The AI doesn’t want them.
It prefers a small number of highly structured pages over hundreds of empty shells.
I observe with my clients that stopping production in favor of consolidation delivers faster results.
We stopped writing.
We restructured.
We invested in architecture.
Result: +290 % mentions in 12 weeks.
Work on thematic authority pays off.
AI Overviews won’t disappear.
They’ll become standard.
The longer you wait, the wider the gap with your structured competitors grows.
It’s never too late to shift to a semantic logic.
A well-built local card becomes a citation capture machine.
And you—are your local product cards already being cited by AI Overviews?
Audit one of your local pages live with me
In 45 minutes, I’ll review 5 local pages with you and show you exactly what’s missing for the AI Overview to cite them. Real-world, live.
Book a strategic call — 45 minFrequently Asked Questions
What exactly is a local AI Overview?
Google generates a block with its AI. It answers a query. It sits above the Local Pack. It synthesizes info from product cards, local pages, or articles. It can include links to sources.
How do I know if my site is cited in an AI Overview?
I manually monitor 10–15 target queries in private browsing. I use a SERP tracking tool that captures AI Overviews, like SE Ranking or Advanced Web Ranking. When my site is cited, the source link appears in the block.
Which structured data schemas should I prioritize?
For a local e-commerce card, I use LocalBusiness with areaServed, FAQ, and Product if the page shows articles. HowTo schema works well when you detail a process (installation, setup).
How long before my pages are cited in the AI Overview?
After restructuring, with my clients, first citations arrive between 4 and 8 weeks. The timeframe depends on Google’s crawl frequency and the thematic authority you’ve built.
Is the Local Pack still useful?
Yes, especially for direct transactional queries. But on its own, it’s no longer enough: it captures less than 40 % of local searches. You also need to aim for AI Overview citation.
