# How to Get Recommended by AI Agents Like Instinct

> To get recommended by AI agents like Instinct, a brand has to be easy for the agent to find, verify and buy from. That means being named in the sources the underlying language models trust, publishing clean product and pricing data in feeds agents can read, and earning reviews and mentions that confirm what you claim.

By Othmane Khadri, founder of Earleads. Published 2026-10-05. Canonical: https://www.earleads.com/blog/how-to-get-recommended-by-ai-agents/

## What is Instinct, and why does it change how brands get picked?

Most advice on how to get recommended by AI was written for chatbots that answer a question and stop. Instinct goes further. It's an AI agent built by Spear Street Technology in San Francisco and founded by Noah Shinn, a former research scientist at Sierra. You text it over iMessage or WhatsApp, or you call it, and according to [Instinct's own site](https://instinct.com/) it "uses its devices in the same way that humans do," with access to your email, messaging, screen, audio and location.

That access changes the job. A chatbot hands you a list of names and you do the rest. Instinct takes the whole errand, so it picks the restaurant, the hotel or the pair of shoes and then goes and books or buys it.

The growth explains why brands are paying attention to Instinct AI now. Instinct opened a private beta in February 2026, moved to an invite only beta in August 2026 with five invites per user, spent $0 on marketing, and still saw invites resold on eBay, as [Fortune reported on September 30, 2026](https://fortune.com/2026/09/30/noah-shinn-instinct-ai-assistant-meta-muse-alexandr-wang-tech-series-c-ai-agent-mark-zuckerberg/). It raised a $250 million Series B at a $2.5 billion valuation from Index and Benchmark in August, then a [$1 billion Series C at a $10 billion valuation](https://techcrunch.com/2026/09/28/viral-ai-agent-instinct-raises-1b-series-c-at-a-10b-valuation/) from Sequoia, Benchmark and Coatue on September 28, 2026. Shinn has said the platform handles roughly $1 billion in annual transaction volume, that [more than 50% of transactions are travel related](https://techcrunch.com/2026/09/29/instinct-founder-said-more-than-50-of-transactions-on-the-platform-are-travel-related/), and that the team is 14 people.

Fourteen people moving about a billion dollars a year don't run a partnerships desk that takes calls from brands. Whatever Instinct picks, it picks from what it can read, check and pay for. That's the lens we use for the rest of this guide.

## How does an AI agent like Instinct decide what to recommend or buy?

Instinct hasn't published how it ranks options. What's public points to three inputs, plus a short list of open questions that any honest guide to agentic commerce should flag rather than paper over.

### The language models underneath Instinct

Instinct's [privacy policy, revised August 26, 2026](https://instinct.com/privacy-policy), says it shares data with "third party AI model providers" without naming them. So Instinct runs at least partly on outside language models. Which model, and which search engine it uses to look things up, is unconfirmed.

Reviewers such as Vellum and Sacra describe Instinct running a persistent cloud computer with a browser and cached logins. Instinct hasn't confirmed that setup, so treat it as a description, not a spec.

The practical point holds either way. If an outside model shapes Instinct's shortlist, then what that model already believes about your category travels into the agent. A brand that's missing from ChatGPT, Gemini or Perplexity answers today has no good reason to expect better treatment inside an agent built on similar technology.

### Merchant feeds and catalogs Instinct can search

The second input is structured product data. Instinct's founder announced a tie up with Shopify in late September 2026, saying "Instinct can now search Shopify's merchant network for products, and find live information on prices, sizes and availability... check out with Shop Pay," as covered by [PYMNTS](https://www.pymnts.com/news/artificial-intelligence/2026/shopify-opens-store-checkouts-to-ai-agents/). Shopify's own pages don't name Instinct yet, so the announcement comes from Instinct's side.

The same week, [TechCrunch reported](https://techcrunch.com/2026/09/28/shopify-opens-checkout-to-browser-based-ai-agents/) that Shopify opened checkout to browser based AI agents through WebMCP, and that some large retailers block agents. For a Shopify merchant, that makes catalog quality the most direct lever on whether Instinct can find and buy your product.

### Instinct Selections and curated picks

The third input is curation. Instinct launched Selections around September 29 and 30, 2026. Selections pushes picks in dining, travel and shopping that users didn't ask for. Instinct describes them as curated by humans (chefs, designers and travel guides) and informed by the user's email and trips, according to [TechCrunch's coverage of the launch](https://techcrunch.com/2026/09/30/instincts-new-product-recommendations-are-giving-some-users-the-ick/). Some users pushed back on the unprompted picks.

How Selections makes money, if it does, is unconfirmed. For a brand, the takeaway is that Instinct's curators are people who read reviews, menus, press and guides, so the earned coverage that wins a chef or a travel guide over also feeds Instinct.

### What we don't know yet about Instinct

A lot, and it's better to say so plainly:

- Which language model or models Instinct uses, and which search engine it queries.
- Whether Instinct will offer any paid placement or merchant program. None is public today.
- Whether Selections is monetized.
- How its terms will settle. In August 2026 Instinct faced [privacy and security concerns](https://techcrunch.com/2026/08/24/instincts-powerful-ai-assistant-is-raising-privacy-and-security-concerns/) over broad license terms and screen and keystroke collection, and rewrote its terms on August 26.

Anyone selling a guaranteed "Instinct ranking" right now is selling something nobody can measure.

## Mentioned in an answer vs picked by an AI agent

Most guides on AI product recommendations treat every AI surface the same. They aren't. Being named in a ChatGPT answer and being picked by an agent that books or buys are two different outcomes, with different ways to lose.

| | Mentioned in an AI answer | Picked by an AI agent like Instinct |
|---|---|---|
| What the user gets | A short list of names to check | A booked table, a paid order, a confirmed trip |
| What decides it | Sources the model trusts, clear pages | The same sources, plus live price, size and availability data |
| Where you drop out | You aren't in the sources | You aren't in the sources, your data is stale, or checkout blocks the agent |
| Who visits your site | Often the human, after the answer | Possibly no human at all |
| Named ranking signals | Not published by the AI labs | Shopify lists data quality, relevance, availability, pricing and engagement signals |

Our read is simple. A mention is the entry ticket and the pick is the sale. An AI agent buying on someone's behalf has extra reasons to skip you that a chatbot never had, because if it can't confirm the price, find the right size or finish the checkout, the next brand on the list is one step away. So we'd rank the work in the order below, and we wouldn't skip the boring data steps for more content.

## Step 1. Get named in the sources the language models trust

Getting recommended by AI agents starts upstream, in the language models and search results they draw on. Since Instinct's privacy policy confirms it uses third party AI model providers, your standing in mainstream AI answers is the closest proxy you have for your standing inside Instinct.

If you've been asking how to get cited by AI, the answer is mostly about agreement across independent sources. A model is more confident naming a brand when your site, your reviews, press coverage, comparison articles and community threads all describe you the same way. In practice that means:

- One plain sentence that says what you sell, to whom, and at what price point, repeated word for word across your homepage, your about page and your profiles.
- Reviews on the platforms your buyers already trust, whether that's Google, Tripadvisor, G2 or a niche marketplace.
- Mentions in the roundups and comparison articles that rank for your category today.
- Real answers in the community threads and forums where people ask for recommendations in your space.

The same work answers the question of how to get ChatGPT to recommend your business. We'd argue it also matters for Instinct Selections, since chefs, designers and travel guides read the same press and reviews everyone else does.

The most common gap is a brand that describes itself four different ways across four channels. An AI agent reading all four can't tell which one is true, and a confused model tends to recommend someone else.

## Step 2. Make product and pricing data machine readable for AI agents

This is the step that separates being mentioned from being bought, and it's the one most brands skip. AI shopping agents need structured data they can trust at the moment of purchase, and an AI shopping agent with a wrong price in hand will buy the wrong thing or nothing at all.

The public signals are clear. Shopify's [announcement on agent shopping momentum](https://www.shopify.com/news/agentic-commerce-momentum) says its Agentic Storefronts push Shopify Catalog data to ChatGPT, Microsoft Copilot, Google's AI Mode and Gemini. Brands not on Shopify can join through what Shopify calls the "Agentic plan." Google's Universal Commerce Protocol (UCP), which sits behind much of this, is backed by Walmart, Target, Etsy, Stripe, Visa and Mastercard.

The volume is real too. Shopify reports that [AI driven traffic to Shopify stores grew 8x year over year in Q1 2026](https://www.shopify.com/blog/how-agentic-commerce-works), and that orders from AI searches grew nearly 13x. In the same piece Shopify names the factors that decide which products surface, namely "data quality, relevance, availability, pricing, and engagement signals." Four of those five are things you control in a feed.

Where to put your data:

1. Shopify merchants should turn on Agentic Storefronts and clean the catalog before anything else.
2. Brands on other platforms should onboard through [Google Merchant Center for UCP](https://developers.google.com/merchant/ucp), which powers agentic checkout in AI Mode and Gemini.
3. Everyone selling products should join [Perplexity's Merchant Program](https://www.perplexity.ai/hub/blog/shop-like-a-pro). It's free, and Perplexity says it gives an "increased chance of being a recommended product."

What "clean" means in practice: every variant has its own price, size and stock status, prices match between your feed and your checkout, titles say what the product is in the words buyers use, and out of stock items are marked as out of stock instead of left live. Travel, dining and service brands have the same job with rates, hours and availability.

If you only fix one thing this quarter, fix the feed. A blog post can earn you a mention. A wrong price in your catalog can cost you the sale after the agent has already picked you.

## Step 3. Remove friction at checkout and booking

An AI agent like Instinct tries to finish the job after it recommends, which means your checkout and booking flow is now part of how you get recommended.

The rails are being built in public:

- Shopify opened checkout to browser based AI agents through WebMCP, and Instinct's founder says Instinct can check out with Shop Pay.
- OpenAI's Instant Checkout in ChatGPT runs on the [Agentic Commerce Protocol, developed with and open sourced by Stripe](https://stripe.com/newsroom/news/stripe-openai-instant-checkout). The [protocol's own site](https://www.agenticcommerce.dev/) documents the spec. Shopify says ChatGPT buyers complete the purchase on the merchant's store in an in app browser, so the exact flow varies by platform.
- Google's UCP powers agentic checkout in AI Mode and Gemini, and the merchant stays the merchant of record.
- Perplexity offers Instant Buy with PayPal.

TechCrunch also reported that some large retailers block AI agents. That's a legitimate choice, but treat it as a business decision with a cost, not a default your security team set years ago. If a bot rule, a CAPTCHA wall or a forced account signup stops an agent halfway through, the agent completes the purchase somewhere else.

Travel brands should look hardest here. More than half of Instinct's transactions are travel related, according to its founder, so hotels, airlines, tour operators and booking sites are where agent checkout already matters most. Walk your own booking flow as if you were a browser with no patience. Every extra step is a place an agent can stall.

## Step 4. Publish answer first pages for the prompts buyers delegate to AI agents

People hand Instinct errands, like "book dinner for four near the office on Thursday, somewhere quiet," or "find trail running shoes in my size that arrive before the weekend." The page that wins is the one that answers the constraints in that errand directly.

That's the core of AI agent optimization at the content level. For each errand your buyers delegate, publish a page where the first paragraph answers it, with the specifics an agent needs to verify a match:

- Price and what's included, stated in numbers.
- Who it suits, and who should buy something else.
- Availability, delivery windows, booking rules or opening hours.
- Comparisons against the obvious alternatives, with real criteria.
- Your brand name in each section, so a passage still makes sense when an AI pulls it out on its own.

Write each section so it could be lifted alone and still name the product, the number and the condition. AI answers quote passages, not whole pages, and a paragraph that says "it's great for this" loses its meaning once the heading around it is gone.

We'd also avoid writing for Instinct by name in your product copy. You don't know which model reads the page, so write for the buyer's errand and let every AI agent read the same clear answer.

## What changes for B2B brands when AI agents do the shortlisting?

Instinct's public numbers lean consumer, with travel above half of transactions and Shopify as the headline partner. B2B brands shouldn't read that as a pass. The same mechanics apply when a buyer asks an AI assistant to shortlist vendors, compare pricing tiers or draft a recommendation for their boss.

For B2B, the "feed" is your public site. An AI agent can only verify what it can read, so the B2B equivalents of clean product data are:

- A public pricing page, or at least published starting prices and the factors that change them.
- An integrations page listing the tools you actually connect to.
- Security and compliance pages a procurement step can check.
- Comparison pages that honestly say who should pick a competitor instead.
- Reviews on G2 or the directories your category's buyers use, written in language that matches how you describe yourself.

Our take is that B2B brands hiding pricing behind a demo form will feel this first. An agent asked for "three CRM tools with per seat pricing inside our budget" can't include a vendor whose price it can't see.

This is the work we do at Earleads as an [AI search visibility agency for B2B and B2C brands](https://www.earleads.com/), and the playbook is the same on both sides. If you're looking for an agency that can get your brand recommended by Instinct, ask how they'll improve your sources, your data and your checkout access, and walk away from anyone promising a guaranteed placement.

## How do you measure whether AI agents recommend your brand?

You can't log into an Instinct dashboard to see where you rank, because no public merchant dashboard exists. Measuring whether AI agents recommend you means testing the surfaces you can reach and watching the traffic that comes back.

A measurement routine that works without any private access:

1. Write a fixed list of errands your buyers would actually delegate, in their words, with real constraints like budget, location and timing.
2. Run each one monthly in ChatGPT, Gemini, Google's AI Mode, Perplexity and Microsoft Copilot. If you have an Instinct invite, run them there too.
3. Log four outcomes per errand, meaning whether you were mentioned, recommended first, linked, and purchasable inside the AI surface.
4. Check feed health in Shopify and Google Merchant Center, since disapproved or stale products can't be picked.
5. Track AI referral sessions and orders in your analytics, so you can see whether Shopify's 8x traffic trend is showing up for you or passing you by.

Keep the errand list fixed so month over month changes mean something. If you want an outside view of where you stand across sources, feeds and checkout, our free [AI visibility audit](/#audit) checks if AI agents recommend your brand today and shows which of the four steps above to fix first.

## Frequently asked questions

### Can you pay Instinct to recommend your brand?

Not through any public program. Instinct hasn't published a merchant program or an ads product, and it describes its Selections feature as curated by humans such as chefs, designers and travel guides. Whether Selections will ever make money from brands is unconfirmed. Today the only route to an Instinct recommendation is being findable, verifiable and easy to buy from.

### How do I get my products into AI shopping agents like Instinct?

Start with the catalogs agents already read. Instinct's founder announced that Instinct can search Shopify's merchant network for live prices, sizes and availability and check out with Shop Pay. Shopify merchants should keep that catalog clean. Other brands can use Shopify's Agentic plan, Google Merchant Center for AI Mode and Gemini, and Perplexity's free Merchant Program.

### How do I get ChatGPT to recommend my business?

ChatGPT recommends businesses it can describe with confidence, so the work is making that description easy. Publish pages that answer buyer questions directly, get mentioned in reviews, press and community threads that agree with your own site, and keep pricing public. Merchants can also reach ChatGPT through Shopify's Agentic Storefronts or Instant Checkout, which runs on Stripe's Agentic Commerce Protocol.

### Does an AI agent optimization agency guarantee recommendations?

No, and you should walk away from any agency that promises one. Nobody outside OpenAI, Google, Perplexity or Instinct controls what their models and agents pick. A good AI agent optimization agency improves the inputs those systems use, meaning sources, product data and checkout access, then measures whether recommendations actually move across a fixed set of buyer prompts.

### How long does it take to get cited by AI?

It depends on which input you change. Product feed fixes reach shopping surfaces as fast as the catalog syncs. New pages can show up in AI answers that search the live web once they're indexed. Changing what a language model already believes about your brand takes longest, because it depends on many independent sources repeating the same facts about you.

### Is getting recommended by AI agents only for ecommerce, or does it work for B2B?

It works for both. Instinct's public numbers lean consumer, with more than half of its transactions in travel. But the same mechanics apply when a buyer asks an assistant for a vendor shortlist. B2B brands need public pricing, integration and security pages, plus consistent third party coverage, so an agent can verify them the way it verifies a product.