How to Get AI to Recommend Your Business
Personal AI agents like Muse, Instinct, and GrokBot are now the gatekeepers between your brand and your buyer. Here's what's actually happening, why your current marketing playbook wasn't built for it, and what to do differently.
Create, optimize, and measure content that actually drives leads with Content Hub.
AI agents are beginning to handle the entire buying process on behalf of consumers, and the early numbers suggest it's moving faster than most marketing teams expected.
Tools like Meta's Muse and Instinct are personal AI agents that connect to your calendar, email, and payment accounts to research, shop, and make purchases on your behalf. Muse hit 3.4 million downloads in under three weeks. Instinct is already processing $1 billion in gross merchandise value.
The agents your consumers have delegated buying to don't respond to clever positioning or warm brand storytelling. They respond to proof, transparency, and validation. This means your campaigns built to reach humans were built for a different era.
What Kipp and Kieran learned testing Muse, Instinct, and GrokBot
Kieran mapped out how marketing has shifted across three eras.
-
In the direct era, brands went straight to consumers — nothing in between.
-
In the intent era, algorithms filtered for intent and you competed to get filtered in.
-
In the delegation era, the consumer hands the entire research and buying process to an agent. You don't get filtered in or out of a results page — you get removed from the process entirely unless the agent decides you're worth surfacing.
Getting filtered out of Google still means you exist. Getting ignored by an agent means you were never part of the conversation.
This isn't a future scenario. Stripe published data showing agent purchases through its Link product are up 38x. Kieran demonstrated buying sneakers through Instinct on WhatsApp — the agent found the product, checked for discounts, and was ready to complete the purchase without him visiting a single brand's website. Kipp used Muse to identify marketing leaders who engaged with Marketing Against the Grain’s Instagram content and enrich those contacts with public email and phone data. The actions agents can take are already well beyond search and discovery.
The last 10–15 years of B2B marketing were built around capturing demand — intercepting high-intent buyers with webinars, ebooks, paid search, and gated content. That playbook assumed a human was at the keyboard doing the research. Kipp's assessment: if you're still playing at the bottom of the funnel trying to capture demand, you're going to get hit hard. You have to move further up — to brand awareness — so agents even know who you are before the buying decision begins.
Agents act like engineers evaluating a purchase decision. They validate claims against external evidence. They don't care about your tagline, your brand story, or your carefully crafted positioning — all of that gets rewritten in the agent's own language anyway. What agents respond to: transparent pricing, specific case studies with measurable outcomes, third-party validation, and public customer reviews. Hidden pricing and vague claims don't create curiosity in an agent the way they might in a human. The agent moves on to whoever gives it what it needs.
Agents can continuously compare you against alternatives, monitor for poor service, and actively look for better options. They can migrate your customer's data to make switching near-zero friction. The old lifecycle marketing playbook of keeping customers through inertia no longer works when an agent is actively auditing their stack on their behalf. You have to win on actual product value and usage.
The one signal agents consistently respond to is consensus. What are real customers saying publicly? Kipp's framing: investing in customer love and turning customers into advocates has always mattered — in the agent era it's 10x more important. Agents are sophisticated peer review engines. If your customers aren't talking about you positively in public places agents can find, you have no signal for agents to surface.
What to do differently: three areas to act on now
1: Agent discoverability
Agent discoverability is how well AI systems can retrieve accurate, comprehensive information about your brand when generating a recommendation.
The important distinction from traditional SEO: you're not optimizing for clicks. The agent may never send a user to your website. It might just include you in a shortlist with a two-sentence description it generated from what it knows about you. That description — a representation of your brand in a room where you weren't present — is what you're optimizing for.
Audit this layer in 30 minutes:
Everything written about your brand up to the model's knowledge cutoff: press, reviews, forum discussions, analyst reports, LinkedIn content from your executives, podcast transcripts
Agents with browsing active (Perplexity, ChatGPT) pull real-time content from your blog, G2 profile, review sites, and linked third-party sources
Schema markup, Wikidata entries, Google's Knowledge Panel, and similar machine-readable representations of your brand identity
Analyst reports, industry rankings, named publications. Being cited in a Gartner Magic Quadrant matters to an agent for the same reason it matters to a human buyer: categorical legitimacy
Open ChatGPT, Perplexity, Claude, and Gemini. Ask each one to describe your company, explain your positioning, and list your main competitors. Compare what they say to how you'd actually describe yourself. The gaps are your content and PR gaps — and they're gaps that agents are filling in with whatever they can find.
Also check: Does your brand have a Wikidata entry? Is your Google Knowledge Panel accurate? Do you have consistent schema markup across your site? These are machine-readable signals almost no one is maintaining, and they directly influence how agents represent you.
2. Agent trust signals
Once an agent can find you, it has to decide whether to recommend you. Trust signals are the indicators that tell an agent your brand is a credible answer to this buyer's question.
The five signals that matter most:
G2, Capterra, Trustpilot, and category-specific review platforms are heavily weighted by retrieval-augmented agents. Volume matters, but recency matters more — a cluster of reviews from four years ago is a liability. You need a consistent review generation motion built into your CS and onboarding workflows, not a one-time sprint.
When third-party experts — analysts, journalists, practitioners — write about your category, is your brand cited? Being quoted in practitioner communities (Slack groups, subreddits, LinkedIn) by people who are themselves credible adds a peer-validation layer that agents interpret as social proof at scale.
An agent can't evaluate vague testimonials. Specific, named, verifiable customer outcomes — "Company X reduced sales cycle from 47 days to 31 days" — are the inputs agents can actually synthesize into a recommendation. Anonymized case studies are nearly worthless here. Named, specific, verifiable outcomes are trust infrastructure.
Agents doing category research evaluate the quality of content in that space. Comprehensive, practitioner-quality content on the problems your category solves registers as authority. Thin blog posts don't. Write the resource an agent would cite, not the post that ranks for a keyword.
Is there cross-source agreement that your brand belongs in your category? If G2 puts you in one category, your website in another, and analysts in a third, agents will hedge or deprioritize you. Coherent, repeated, cross-source classification is trust infrastructure.
3. Brand awareness
The playbook built around capturing demand worked when humans were doing the research and needed to be intercepted at the moment of high intent. Agents compress or eliminate that moment entirely. By the time an agent is actively evaluating vendors, the consideration set is largely already formed by what the agent's training data and real-time research have surfaced.
Getting into that consideration set requires building brand awareness earlier and higher in the process — before the buying decision is triggered.
Create content that gets referenced, not just content that gets traffic. Agents pull from the web, from forums, from review sites, from publications. Content that earns genuine third-party reference builds the kind of presence agents can find and use. A piece that gets cited in five industry publications does more for agent discoverability than a piece that gets 10,000 organic visits.
Invest in executive visibility. LinkedIn content, podcast appearances, and bylines from your founders and senior leaders seed your brand's perspective across training corpora at a human level. Agents retrieve and weight named expert perspectives differently from anonymous company content.
Build your customer community into a public signal. Customers talking about you in places agents can access — forums, review sites, LinkedIn posts, Reddit threads — is more valuable than almost any owned marketing asset in this era. Customer advocacy programs and community building have always had a strong ROI case. It just got significantly stronger.
4. Conversion architecture
This is the least-developed layer across almost every brand — and where the real competitive advantage will be built over the next 18 months.
Conversion architecture is the infrastructure that allows an agent to not just recommend your brand, but to initiate, facilitate, or complete a purchase action on behalf of the user. It's the difference between being on a shortlist and closing the deal, even when no human is clicking through your funnel.
What you can do today:
Structured schema markup (schema.org/Product, schema.org/SoftwareApplication) that defines pricing, features, use cases, and integrations in a format agents can parse. If your product information is buried in marketing copy, agents will skip to a competitor whose data is cleaner.
Some brands are already publishing pages that explicitly address what an AI agent needs to represent them accurately — clear product descriptions, pricing structures, integration capabilities, current customer use cases, and updated trust signals in a single structured document. Think of it as a media kit for the AI layer.
Agent-referred traffic currently looks like direct or dark social. Start tagging your key conversion pages and watch for patterns when you optimize your AI-facing content. As agents gain the ability to take direct action — book a demo, start a trial, request a quote — your conversion pages need to be structured for machine-initiated sessions, not just human ones.
Kieran said it directly in the episode: agents hate hidden pricing. Pricing cannot be friction. If a buyer agent encounters a "talk to sales" wall where pricing should be, it moves on to whoever is transparent. This isn't just a conversion tactic — it's a prerequisite for being considered at all.
Checklist
The Agent-Era Brand Audit
Run these questions against your website and current marketing assets.