Brand Discovery is Being Reshaped by AI: Generative Engine Marketing and the Future of Influence

The era of marketing to humans who make purchase decisions is changing dramatically. Humans still choose brands and products, emotions still matter, brand still matters. But the fundamental shift is that the path to that decision looks very different today. 

A runner who needs shoes used to search, scan links, compare reviews, visit brand sites, and build their own shortlist of which new shoe they wanted. Now they ask an AI assistant. 

The assistant scans hundreds of sources in seconds, synthesizes expert reviews, Reddit threads, retailer pricing, and brand content, then presents a tight shortlist. The person is still choosing from the list, but only from what the models are surfacing. 

So, the implication for brands is that if you’re not in the model’s consideration set, you’re unlikely to be in the humans. 

Ultimately, your brand narrative, your technology story, your competitive positioning; it's being written by people who don't work for your brand, and then repeated at scale by AI agents.

This has restructured how brands earn visibility, authority, and preference. The medium that delivers your message isn’t in brands’ full control. It’s shaped by third parties, before it even reaches a potential buyer. 

This is the era of Generative Engine Marketing. 

How we tested the shift

Using Jellyfish Agent Shopper platform, we simulated 50 structured shopping tasks within a single category: running shoes. We did 50, because we wanted coverage across the most common intent states within one purchase journey rather than surface level volume, for the purpose of this report. The tasks were structured across five distinct intent clusters:

• Category entry: ‘I need new running shoes’
• Performance: ‘Best shoes for marathon training’
• Injury or condition: ‘I have shin splints, what should I buy’
• Value sensitivity: ‘Bust cushioned running shoes for under $120’
• Replacement decision: ‘Should I replace my current shoe’

Each task was standardized and submitted through Agent Shopper across multiple leading LLM environments. For every task, three coordinated agents per model analyzes the market, compared products and pricing, and returned a structured report detailing:

• Products recommended
• Attributes emphasized 
• Sources cited
• Retail routes suggested

The goal was to understand how AI agents evaluate, frame and route customer decisions and what patterns emerge. 

The results reveal how marketing power is being redistributed. 

What AI Shopping Agents revealed and what it means for brands

Your AI brand identity is narrower than you think

One major brand appeared in 70% of all shopping tasks. That sounds really great and like they’re dominating. Except agents consistently recommend only two of the brand’s eight core models. And they frame the brand the same way every time: ‘great cushioning.’

Not about speed, trail, innovation. Just cushioning. 

AI agents learn from the most reinforced signals in the ecosystem. If cushioning is your most reviewed attribute, that becomes your AI identity. 

So, your brand positioning may be expansive. But your AI positioning may not be. 

Why this matters

If your brand identity is defined by your most reviewed attribute, your growth ceiling is constrained by past conversation, not future strategy. 

So, what should you do?

• Audit how LLMs currently describe your brand across key demand scenarios
• Identify which attributes dominate and which are invisible
• Activate content, PR, creator, and retail ecosystems to rebalance that narrative intentionally. 

If brands don’t shape the attribute mix, the ecosystem will shape it for you. 

Your brand story is being written by people who do not work for you

When agents cited sources for their recommendations, brand websites were referenced primarily as places to transact, not as sources of authority. The authoritative voices shaping recommendations were third parties like publisher reviews, specialist media, retailer content, forums. 

Your brand narrative, technology story, and competitive positioning are being written externally, then repeated at scale by AI models.
Historically, marketing controlled the message and optimised distribution. Now the message is synthesized from multiple sources and weighted by perceived authority. 

Why this matters

In today’s marketing era, you can’t just publish your story. You must ensure it is echoed, validated and structured across the ecosystem models draw from. 

So, what should you do? 

• Prioritize expert validation and category authority, not just reach
• Align PR, creator strategy, affiliate and retail content with core positioning pillars
• Ensure consistency across retailer PDPs, reviews, third party coverage. 

Marketing strategy has to account for how influence compounds inside models. 

AI pricing agents are the most effective discount channel you never approved

When asked to find the best price for a specific product, agents consistently pushed routed shoppers away from direct to consumer channels and toward discount retailers, marketplaces or coupon strategies. 

Agents optimize for value on behalf of the user. They are essentially training customers to expect a deal and bypass your DTC channel entirely. And they do it confidently and instantly. 

In effect, AI assistants are rewiring your distribution strategy, becoming an unapproved discount channel. 

Why this matters

If your pricing, retail and feed strategies are fragmented, AI agents will exploit that fragmentation. 

So, what should you do? 

• Align pricing architecture across DTC and retail patterns
• Strengthen product feed accuracy, availability and schema
• Monitor how agents route transactions and where margin is being lost

Agentic commerce isn’t just about product visibility but marketplace strategy. 

Structured content wins

Across tasks, agents recommended products they could confidently describe. Brands with detailed specifications, structured product data, comparison content, and robust review ecosystems were surfaced more often than other competitors with thinner digital footprints. 

In some cases, category leading products were overlooked simply because there was not enough structured, crawlable information for the agent to synthesize. 

Search optimization helps brands get found. Generative optimization determines whether brands get recommended. 

Why this matters? 

In AI mediated journeys, clarity and structure can outperform creativity alone. 

So, what should you do? 

• Enrich PDPs with structured data and schema
• Build comparison content aligned to real decision criteria
• Ensure technical accessibility for AI crawlers across domains
• Treat metadata architecture as strategically as creative execution

If agents can’t parse it, they can’t recommend it. 

Your biggest competitive risk is model evolution

While this report focused on cross intent patterns, rather than model comparison, one dynamic was clear in the research. As models evolve, how they weight authority, price sensitivity attributes and sources will shift. Your performance inside AI systems isn’t static. It will move as models update. 

Why this matters

Optimization is not a one time audit. It’s an ongoing visibility discipline. 

So, what should you do? 

• Continuously monitor Share of Model visibility 
• Track shifts in attributes emphasis and source citation
• Build multi-format content infrastructure that performs across model types

Brands that build resilient ecosystems that perform across change will stand out above those that only optimize for a single model version. 

AI has changed control, not just discovery

It’s largely understood that AI has changed discovery, and it’s true what we say about brand influence decisions. But what is different now is scale and synthesis. AI systems compress third party opinion, pricing data, expert validation, and brand messaging into a single answer. That answer often replaces browsing journey. 

It’s not that marketing controls the medium end-to-end, but the medium itself synthesizes and prioritizes signals. That shifts the weighting across creator, retail, SEO, feed management and technical architecture. This is restructuring how brand earn consideration. 

So, now I leave you with the new imperative - build for the model AND the human. 

It’s all about measuring how AI agents discover and frame your brand, engineering structured, machine readable content ecosystems, building third party authority intentionally and protecting transaction routes and margins. This is your playbook for how your brand will be recommended.