How to get AI Overview & AI Mode to recommend your products | A Complete Guide

over the past year, our team has been running a bunch of tests on how to rank on ai overview and ai mode.

we've figured out a few reliable methods to get our clients' products showing up in these recommendations

and we're seeing solid traffic & sales come from them.

now in case you’re unfamiliar with ai overview and ai mode

ai overview is the ai-generated summary that appears at the top of many searches.

it gives you a quick answer and links to relevant websites

ai mode is a separate ai-powered search experience inside google.

you ask questions, get an ai-generated answer, and keep asking follow-up questions like chatgpt.

imo, the majority of brands are underestimating how big ai mode and ai overview are right now.

i was talking to one of our partners recently and they made an interesting point.

a while back, chatgpt used to drive most of the llm visibility and clicks for ecommerce brands.

but now…

it's getting close to an even split with google's ai mode and ai overview.

how many posts or youtube videos have you seen about getting your brand into chatgpt?

now compare that to how much you've seen about ai overview and ai mode.

probably not much.

and yet, those two placements are taking ai-driven traffic away from chatgpt.

honestly, i wouldn’t be surprised if they surpass chatgpt at some point

let me explain why…

why google's ai is pulling ahead of chatgpt

1. google is putting ai into products people already use every day

ai overview and ai mode sit right inside google search.

billions of people have been using google for decades.

there's almost nothing new to learn.

people search the same way they always have, just with ai built into the experience.

to get people buying through these ai experiences...

google didn't need to attract new users.

they just needed existing users to get comfortable with a different way of searching.

meanwhile, chatgpt is still growing its user base.

chatgpt got there first when it came to ai product recommendations.

but google has one advantage that chatgpt simply can't, and probably never will, replicate.

2. people are getting more comfortable using ai inside google search

google has improved ai overview and ai mode a lot over time.

as gemini, the model behind them, keeps improving...

the answers become more accurate, more useful, and more reliable.

when the experience gets better, people use it more.

on top of that...

google is giving these placements a lot more visibility.

for years, the first thing you'd see on a search page was a search result or a shopping carousel.

now it's often an ai overview.

it takes up almost half the screen before you even get to the normal results.

we're also seeing more searches where people never click a traditional search result.

they read the ai response, trust it, and in a lot of cases, make a purchase from there.

3. google is THE go-to platform for product research

people have been researching products on google for decades.

it's where they compare brands, read reviews, and decide what to buy.

that habit is hard to change.

early on, chatgpt had a clear advantage.

you could ask long, detailed questions with plenty of context and get great recommendations.

at the time, it felt like a completely new way to find the right product.

but google can now do the same thing.

you can ask those same questions without leaving google search.

so that advantage isn't really there anymore.

and i don't see a strong reason for most people to change where they do that research.

when i'm researching a product, google still feels like the most natural place to do it.

4. google is in a strong position to win the ai commerce race

chatgpt knows what you've shared in a conversation.

google knows what you've searched for, and more importantly, what you've bought.

they've built decades of shopping signals through search and shopping, backed by billions of searches every day.

when you're trying to predict what someone is likely to buy...

that's a much stronger foundation.

commerce is also not chatgpt's main focus.

for google, helping people discover and buy products has always been a core part of the business.

and because advertising is such a huge business for them...

they have every reason to keep investing in ai-powered shopping.

you can already see it in how quickly they're shipping updates.

for example...

in some accounts, there’s now a dedicated generative ai section in google search console.

now, i know this probably sounds biased coming from someone who runs a google ads agency.

but i'm not trying to take shots at chatgpt here.

it's absolutely a real acquisition channel.

my point is:

google is becoming a serious player in this space, and i still don't think enough brands are paying attention.

in my opinion...

as google keeps improving ai overview and ai mode, and people get more comfortable buying through them...

brands that move early are going to have a real advantage.

i think these placements will become a meaningful revenue channel alongside traditional google ads campaigns.

with that being said…

how do you get your products recommended by ai mode and ai overview?

first, you need to understand that the fundamentals of traditional seo are still the backbone of how ai ranking systems work.

ai models pull from authoritative sources and surface brands with strong signals across the web.

they trust the same things google trusts:

• quality content

• structured information

• credible mentions

• strong review profiles

so if your brand isn’t visible in ai answers, it comes down to 2 core problems.

why your products don’t get recommended by ai overview and ai mode

1. there simply isn't enough information about you for ai models to find

these models build their understanding of the world by crawling enormous amounts of content across the web.

when someone asks "what's a good led face mask for anti-aging"...

the ai draws from everything: articles, reviews, forum threads, product pages, reddit discussions.

if your brand has a thin footprint across those sources, the ai either doesn't know you exist or doesn't have enough to confidently recommend you.

2. the information that does exist is hard for ai to parse

ai builds its understanding of your brand from content that exists across the web.

the problem is that a lot of brand information online wasn't written for machines to interpret.

your own website might have information buried inside an image, inside a pdf, or written in a way that requires context from multiple sections to fully understand.

plus, a lot of what's out there is just inaccurate.

wrong pricing, outdated specs, details from a product version that no longer exists.

the result is that brands get misrepresented, underrepresented, or just skipped entirely.

this is the starting point for understanding how ranking on ai chatbots works

it's about whether your brand is present enough and accurate enough across the web for an ai system to confidently surface you.

now that you understand that, let's get into the tactical strategies that've been working well for us.

how to show up on ai overview & ai mode

1. optimize your existing content for ai models

the first place to start is making sure your current pages are readable by ai.

ai models pull responses by finding text that cleanly and immediately answers a specific question.

if your content buries the answer deep in your site, or requires context from elsewhere on the page to make sense, it gets passed over in favor of a source that just states the answer upfront.

the fix comes from understanding what people are asking ai mode and ai overview about your product category.

then structure your content to answer those questions

if people are asking "what's the most affordable [product]," you need a section that addresses your price and how it compares to competitors

if they're asking "is [product] comfortable to wear," you need sections that answer those questions

schema markup is the other lever here.

it adds a small piece of extra code to your page (which is readable by ai) that explicitly labels your content.

2 types that every brand should have.

> faq schema, which labels your question-and-answer content so ai can pull it when someone asks a relevant question.

> product-specific schema, which labels your product name, price, availability, and specifications.

2. build new pages designed for ai models

next, you want new pages around the questions people are asking before they buy.

these pages aren't built for google's search results. most visitors will never land on them.

their only job is to give ai a clean, structured source to pull from.

what pages you build depends on your category and what drives buying decisions for your customers.

take supplement or skincare brands.

trust and certification are major factors, and FDA approval is something a lot of buyers look for.

but the fda website typically groups products under broad categories alongside dozens of other brands, making it hard for an ai model to surface your specific product.

building a dedicated page that contains every relevant certification detail, the classification number, etc… gives the llm a clean source to pull from instead of guessing.

brand fact & faq pages are also worth building.

a lot of brands have inconsistent information scattered across the web: outdated claims, old pricing, specs from a previous version.

a brand fact page gives llms a single source of truth.

when those models have structured information about you, they’re much more likely to recommend your products

faq pages are pages on your site that list common questions and answers about your product and brand

keep answers neutral, specific, and grounded in real ingredients or specs, 3-5 sentences each

write them the way you'd want an ai to repeat them back to a buyer

• your top products

• what makes this product different

• what certifications it holds

• how it compares to alternatives

• what the actual specs are

• etc

3. build listicle content that gets your products into the consideration set

ai isn't only being asked to explain products anymore.

people are asking it to narrow down their options and tell them what to buy.

that's where listicles, roundups and “best of” pages come in.

think:

"best cooling blankets for hot sleepers"

"best led face masks for sensitive skin"

"7 best running shoes for flat feet"

"best [product] under $100"

these pages give ai a structured source that covers multiple products within a category

AND it gives you an opportunity to explain where your products fit.

the important part is that you can’t just throw together a thin page and call your product the best.

you need to compare the products based on things buyers care about:

• price

• features and specifications

• who each product is best suited for

• key differences

• strengths and limitations

you can also use direct comparison formats like "[product] vs [competitor]" if there's genuine search demand.

but the bigger opportunity is category-level content where your product sits next to the options buyers are already looking at.

this type of content can live on your own site

or you can get it placed on relevant third-party sites through outreach and guest posts.

either way, you're giving ai a piece of content that evaluates products in the way buyers are asking it to.

4. get your brand appearing where ai goes to validate reputation

when someone asks an ai about a product or brand, the ai looks for validation from independent third-party sources.

reviews, testimonials, third-party mentions, etc

these are the signals llms use to decide whether your brand is worth recommending.

a page on your own domain saying your product is the best in its category carries less weight than the same conclusion appearing on an independent site.

reddit is the biggest source here.

when someone asks ai overview and ai mode about a product category or brand comparison, reddit often comes up as one of the top sources google pulls from.

but you want to approach this carefully.

a new account that jumps straight into promotional content can get flagged immediately.

the move is building multiple reddit accounts over time by participating in relevant communities.

comment on discussions, post about topics you know about, engage in threads that have nothing to do with your brand.

once an account has enough history and karma…

you can start showing up in conversations about your product category with brand mentions.

beyond reddit, you should get your brand included in:

• independent review sites (you can build these yourself btw)

• comparison articles on external domains

• industry-specific publications

this gives llms a wider pool of credible sources to pull from when someone asks about your product category.

5. clean up your product feed

your product feed is no longer just a google ads asset.

if your feed has outdated pricing, vague titles, or missing attributes, the ai either skips your product or surfaces it with wrong information.

so make sure every attribute in your feed is accurate and updated:

• titles

• descriptions

• images

• identifiers

• reviews

• other relevant attributes like size, color, etc

google also added something to merchant center that many brands don’t know about

it's called conversational attributes

ai mode, gemini, ai overviews, the agentic commerce stuff google is building. all of it needs structured data to confidently match, compare and recommend your product

if your feed doesn't hand it that data, the ai guesses or skips you

conversational attributes are how you feed the machine directly

the main ones live right now:

• question_and_answer: pre-answered buyer questions, up to 30 per product

• document_link: pdf manuals/spec sheets google can pull faqs from

• related_product: tag accessories, substitutes, often-bought-with items

• item_group_title + variant_option: much cleaner variant grouping

• popularity_rank: a 0-100 score based on your actual internal sales data

how to actually do this:

if you've got a small catalog, supplemental feeds are the easy route.

you add the conversational attributes there and leave your primary feed alone. or you handle it through your feed management tool

but if you want to do this at scale, build it into your shopify or store backend so every product going forward carries this data.

set it up once, it flows through your feed automatically.

and if it's living in your backend, there's a good chance google picks it up from your website directly too (this will likely boost your organic ai ranking too).

prioritize bestsellers and your most question-heavy products first. electronics, apparel and footwear with variants, anything with sizing or compatibility questions

starting with bestsellers gives you another benefit. popularity_rank is a 0-100 score showing how well a product sells against the rest of your catalog, and google uses it to push your strongest products on ai surfaces.

when you focus your enrichment on bestsellers, you're not spreading effort thin, you're stacking the richest data on the products google is most likely to surface anyway

for question_and_answer, use real buyer questions. specs, compatibility, usage, sizing. don't just copy your description or highlights.

for related_product, use the relationship types (accessory, substitute, often_bought_with) and reference by internal id or GTIN

if you want to be first, you can start adding these manually today.

but the better move is setting it up in your store backend now so it scales without you having to touch it again

6. on the google side: test campaigns eligible for ai mode & ai overview

google rolled out product recommendations inside ai mode, and the campaign types most likely to serve ads within that experience are pmax, ai max, shopping.

ai max is a big part of this.

people are starting to search the same way they talk to ai chatbots.

that means longer, more specific, and more detailed questions:

"noise cancelling headphones for open office"

"cooling sheets for hot sleepers that don't pill after washing"

"best shoes for flat feet marathon training under $150"

it becomes almost impossible to build a manual keyword list that covers every variation people type.

a lot of these searches also don't generate enough volume to justify targeting individually.

those are exactly the kinds of searches ai max was built to capture.

instead of relying only on your keyword list, it looks at the meaning behind the search.

it also considers user behavior, previous searches, and browsing signals to predict whether someone is actually looking for your product.

let's say you sell pet carriers.

someone searches:

"do i need to bring something to carry my dog on a plane?"

according to google, ai max understands the intent behind that search and recognizes that the user is likely looking for a pet carrier.

that makes your ad eligible to show.

then it dynamically adapts the headline to fit the context, with something like:

"dog carrier that fits under airplane seats."

we've been running ai max across a number of accounts, and we’re seeing a lot of success with it.

now, for all these campaigns to work well, the inputs you feed google need to be in good shape.

ai max can use final url expansion to send people somewhere other than your pdp if it thinks another page is a better match for the search.

that could be a category page, comparison page, collection page, or another commercial landing page.

in other words, your website becomes part of the targeting.

if those pages don't clearly explain what they offer…

google has a much harder time understanding when to show them.

so all of your landing pages should have strong, descriptive copy.

keep your site structure clean too.

a logical hierarchy makes it easier for google to:

• crawl your site

• understand how pages relate to each other

• send traffic to the page that best matches the user's intent.

the product feed is the most direct one.

clean product titles, accurate specs, strong images, etc. things like that heavily affect how google decides which placements to serve your ads on.

beyond the feed, conversion tracking needs to be accurate.

audience signals matter too.

the customer lists, remarketing audiences, and interest segments you attach to these campaigns can give google a starting point for who to target.

in my opinion, the brands that figure out how to run these campaigns profitably now will have a huge head start

---

so that’s a quick guide on how to get your products recommended on ai mode and ai overview.

of course, whether some of these steps apply to your account depends heavily on your category demand, margins, pricing, and product type.

so before you implement anything, make sure it fits your context.

the ecom landscape has always rewarded brands that moved early on distribution shifts before they became the standard playbook.

google ads in the early 2000s. facebook ads in 2013. tiktok shop in 2022.

every time, the early movers captured disproportionate returns before the channel got crowded and expensive.

the signals right now are pointing to llm chatbots being that next shift.

and if you want to move early, start showing up in these placements and capturing sales from them…

my dms are open.

• Jackson