Friday 15 Podcast

Does AI Recommend Your B2B Brand?

Michael Schenck of Jaipur Living joins Andy and Brian to break down the seven steps that actually make a B2B website visible and recommendable to AI shopping engines.

Friday 15 Podcast · Guest: Michael Schenck, Senior Digital Product Manager, Jaipur Living

Key takeaways

  • 73% of US sales and marketing professionals use two or more disconnected tools to get a complete picture of the customer, and two thirds say this causes teams to miss actions and opportunities.
  • Visits to generative AI engines grew from 2.9 billion in 2022 to a projected 114 billion this year, and Google's AI Overviews now appear in 43% of US searches, up from 16% in 2025.
  • AI-referred shopping traffic converts nearly 50% higher than organic search, and outperforms SEO in 23 of 25 merchant categories by an average of 56%, according to Shopify.
  • Markdown rendering at the edge, llms.txt files, and correctly configured robots.txt rules are foundational technical steps for making a site legible to AI shopping agents.
  • Content built for AI agents still needs to hold up for human readers. Pages that read as garbled AI-answer regurgitation risk both a poor customer experience and penalties under Google's newest anti-spam guidance.

Fragmented tools are still costing sales and marketing teams money

The breaking news segment covered a new eMarketer study on data fragmentation. Seventy three percent of US sales and marketing professionals now use two or more tools and systems to get a complete picture of the customer, and two thirds say disconnected tools cause their teams to miss actions, opportunities, and updates.

Andy connected the problem to his own daily experience managing information across scattered tools. “I feel like I’m overwhelmed with information,” he said, describing a stack of Google Sheets, multiple CRM instances, and Gmail search that returns results the moment a word appears anywhere in an email, whether relevant or not. He sees agentic AI as a genuine fix for this specific problem, one that can retrieve and connect information across systems on request rather than forcing a person to manually search each tool. Brian agreed AI has the potential to help, but pushed back on framing the issue as purely a tools problem. “Some of this is self-inflicted,” he said, pointing out that a sales rep who does not use the CRM their company already provides is choosing the inefficiency, not just suffering from it.

Generative AI traffic has exploded, and it is reshaping how Google itself works

The main topic set up a session Master B2B will run at the B2B Exchange at Shoptalk Fall in Nashville. Andy opened with the scale of the shift: visits to generative AI engines were 2.9 billion in 2022 and are projected to reach 114 billion this year. Google itself is becoming an answer engine in the process. AI Overviews now appear in 43 percent of US Google searches, up from just 16 percent in 2025.

Andy admitted he does not fully understand how Google intends to monetize this shift given its historical ad model, even as the company keeps hitting its financial numbers. Brian raised the question directly to the episode’s guest later on, framing it as one to watch rather than one either host could answer definitively.

AI-referred traffic converts far better than traditional search

Andy shared research the episode’s guest had published showing AI-referred shopping traffic converts at meaningfully higher rates than regular SEO traffic. According to Shopify, AI-referred shopping sessions convert nearly 50 percent higher than organic search, and answer engine optimization traffic, AEO for short, outperforms SEO in 23 of 25 merchant categories by an average of 56 percent. Andy explained the mechanism behind the lift: when a buyer asks an AI engine a question, the engine returns a single answer rather than a list of options, so the traffic that reaches a site afterward is already highly qualified. “That’s why the conversion rate’s higher,” he said. “There’s a new layer in front of all of this stuff, and that layer actually does a pretty good job of qualifying the traffic.”

Michael Schenck on what it actually takes to get recommended

Joining the conversation was Michael Schenck, Senior Digital Product Manager at Jaipur Living, sister company to the largest manufacturer of handmade rugs in India and a B2B site selling primarily to interior designers and boutique retailers, alongside direct-to-consumer marketplace channels. Schenck described a collapsing product discovery funnel: buyers increasingly use agentic engines to do the comparison and brand-consideration work that used to happen across many browser tabs, arriving at a site already sold on the brand and ready to buy. “By the time they hit your site, most of the time now they’ve already decided on your brand,” he said. “It’s a very high intent traffic.”

That creates a specific challenge for visually driven products like rugs, where texture, weight, and feel are hard to convey in words. Schenck’s team is working through how to translate that visual differentiation into structured data an AI agent can actually parse, since agents cannot see or feel a product the way a human shopper can.

Seven steps to become AI-visible, built from real practitioner experience

Schenck walked through steps his team is implementing to make Jaipur Living’s site more legible to AI agents, drawn from a forthcoming paper he plans to present at the B2B Exchange and share through the Master B2B forum. He found no single SEO tool covers everything needed, so his team built each piece individually rather than relying on an all-in-one platform, which also avoids dependency on any one vendor as the landscape changes.

Humans make mistakes, they have incomplete thoughts. You can use AI to fill in these gaps and determine what they really meant to say, instead of what they actually typed.

Michael Schenck, Jaipur Living

The first piece is markdown rendering at the edge, a beta capability from Cloudflare (with Fastly and other CDN providers moving toward similar offerings) that converts dense web pages into lightweight markdown files an AI agent can parse in milliseconds, rather than requiring a script to extract meaning from a PDF or Word document. On top of that, an llms.txt file acts like a site’s elevator pitch: a short summary of what the business does and where key pages live, with an expanded llms-full.txt version pointing to key pages and FAQs. Schenck also flagged robots.txt configuration as a common failure point, since misconfigured bot rules can accidentally block the same crawlers a site is trying to attract. On the content side, he recommended structuring information in clear headings, bullet points, and answer-shaped sections that mirror how a buyer would phrase a question, while explicitly warning against generating pages that are just AI-answer regurgitation with no value to a human reader, a pattern Google’s newest anti-spam guidance now penalizes directly.

Balancing content for bots and humans at the same time

Andy asked Schenck how to reconcile writing for AI agents with writing for people, since the two do not always read the same way. Schenck’s answer was that the balance is the whole challenge. “We want to make sure that this content is valuable to both the agents and humans,” he said. “If a human gets to that page and it’s just like a garbled list of text, that doesn’t help anybody, and that’s going to get you flagged.” He sees this as related to, but not identical to, traditional SEO, describing it less as SEO on steroids and more as a related discipline with its own specific caveats.

Frequently asked questions

How much has generative AI traffic grown?

Visits to generative AI engines grew from 2.9 billion in 2022 to a projected 114 billion this year. Google's AI Overviews, which display AI-generated summaries directly in search results, now appear in 43% of US searches, up from 16% in 2025.

Does AI-referred traffic convert better than traditional search traffic?

Yes. According to Shopify, AI-referred shopping sessions convert nearly 50% higher than organic search on average, and answer engine optimization (AEO) traffic outperforms SEO in 23 of 25 merchant categories by an average of 56%. AI-referred traffic tends to be highly qualified because buyers arrive after an AI engine has already answered their question and narrowed their options.

What is llms.txt and why does it matter for AI visibility?

An llms.txt file is a short summary, similar to a site map, placed at the root of a website that tells AI agents what a business does and where its key pages live. An expanded llms-full.txt version points to additional key pages and FAQs, giving AI agents a fast, structured way to understand a site without parsing every page individually.

How does markdown help a website get recommended by AI agents?

Markdown is a simple text format that breaks content into headings, sections, and bullet points, which AI agents can read directly and quickly. Some content delivery networks now render web pages as markdown at the edge, letting an AI agent parse a page in milliseconds instead of needing extra tools to extract meaning from a formatted web page, PDF, or Word document.

Can content be optimized for AI agents without hurting the experience for human visitors?

It needs to be. Content built purely to feed AI agents, with no value to a human reader, risks both a poor customer experience and penalties under Google's newest anti-spam guidance, which targets pages that are essentially AI-answer regurgitation. The goal is structured, answer-shaped content that works for both audiences at once.

Sources & methodology

  1. eMarketer, sales and marketing tool fragmentation study, 2026
  2. Similarweb, 2026 Generative AI Landscape Report
  3. Shopify, Q1 2026 commerce data on AI-referred shopping traffic
  4. Friday 15 Podcast, Master B2B
Andy Hoar Andy Hoar
Co-Founder, Master B2B

Andy is a Co-Founder of Master B2B, founder of Paradigm B2B and author of the book Bot2Bot: The New Future of B2B Commerce. Andy is one of the leading global authorities on B2B commerce strategy.

Brian Beck Brian Beck
Co-Founder, Master B2B

Brian is a co-founder of Master B2B, Managing Partner of Amazon agency Enceiba, and author of the book "Billion Dollar B2B Ecommerce." Brian has also been C-level digital commerce executive with two decades of experience.