July 23, 2026 | A conversation with Nick Sands, Founder at Command Growth; Rishabh Chatterjee, Co-Founder & CTO at Passionfruit; and Josh Weiss, COO & Head of eCommerce at Bambi Baby

How AI Search Is Rewriting Product Discovery

For a category built on long research cycles, AI search is a perfect storm. Josh Weiss, COO & Head of eCommerce at Bambi Baby, described baby gear shopping as three weeks of multi-tab comparisons between brands like UPPAbaby and Nuna, now collapsing into one ChatGPT conversation where a first-time parent can ask for the best convertible car seat for a small SUV under $400 and get a direct answer. That moves awareness and consideration onto a surface brands cannot see or control, which is why Bambi Baby produces deep, well-structured content on its high-priority products, written so AI models read it cleanly and recommend it outright rather than just cite it.

Brands still hold real control over how AI represents them, according to Rishabh Chatterjee, Co-Founder & CTO at Passionfruit, whose firm works with more than 500 brands, roughly 80 percent of them on Shopify. Every AI-generated answer draws from one of three source types: branded content, third-party content, or social content such as review platforms, UGC, and Reddit. On branded queries, branded content carries the most weight, including blogs and product detail pages, where adding 10 to 15 attributes behind the scenes, invisible to shoppers but readable by search engines, makes a measurable difference. Timing varies: a new mention usually gets picked up within one to four weeks depending on crawl frequency, and moving from rank five to rank one takes not one factor but a citation in a frequently cited source combined with strong on-site content, structured data, and technical SEO.

The business risk is subtle, because a brand can hold its top ranking and still go invisible. Josh cited a Google stat showing zero-click searches rose from 56 percent to 69 percent of all Google searches after AI Overviews launched, so a first-place ranking no longer guarantees a click when the answer already appears on the results page. The traffic that does convert has shifted toward higher intent: Bambi Baby saw a 13 times year-over-year increase in AI-referred orders in the first quarter, at a higher average order value, which for Josh makes visibility inside AI answers a requirement for reaching the highest-intent buyers.

SEO Gets You Ranked, AEO Gets You Quoted, GEO Gets You Recommended

The framework both speakers returned to all hour was Josh's: SEO gets a brand ranked, AEO gets it quoted, and GEO gets it recommended, three pieces of one new funnel brands have never had to build before. Rishabh agreed, adding that while the underlying strategy overlaps heavily across all three, the way success gets measured differs by discipline.

Sequencing them as separate projects is a mistake. Josh called the work "the same vertical," having spent much of the year getting technical SEO and structured data in place before investing further in content, since good content on a weak foundation limits what AI systems can surface. Passionfruit does not sell SEO, AEO, or GEO as standalone products either; instead it pitches one organic strategy built from five verticals: content, on-page work, PDP optimization, off-page efforts, and, increasingly, programmatic pages aimed at the longer-tail terms that query fan-out produces.

If one discipline is underfunded relative to its payoff, both pointed to GEO. Across Passionfruit's client base, AI search brings in roughly 5 to 15 percent of both traffic and revenue, with Google still holding the majority, though software companies the firm works with already see 30 to 40 percent of traffic and signups because their audiences adopted these tools earlier, and Rishabh expects ecommerce to reach 20 to 25 percent within a couple of years. The reason brands drag their feet is measurement: no dashboard shows a brand it is being recommended 30 percent of the time for a given search, which makes budget hard to justify to leadership focused on immediate ROI. Josh added that earning organic mentions works like link building but fits no one's job description, and that asking customers to write reviews rarely works as platforms restrict incentives.

Structuring Content and Sites for AI to Read

The biggest change to content strategy starts before a query is ever answered. Rishabhpointed to query expansion, the step where AI systems break a single search into five to seven sub-queries, which means brands can no longer optimize around one keyword and instead have to cover a topic in full. The data behind that content has changed too: beyond keyword research, brands should study the questions people actually ask AI platforms and mine their own review data, which he called a goldmine of what customers praise, criticize, and ask about. Structured elements and FAQs matter more now, since both feed AEO and AI Overview responses directly.

Depth is the point at Bambi Baby, where content is intentionally long-form and written by copywriters with deep product knowledge rather than AI tools, because Josh believes AI-generated content just regurgitates information from elsewhere. It sits alongside structured data and 10 to 15 product attributes behind each PDP, giving models information shoppers never see. Video carries real weight in this mix: Bambi Baby invests heavily in YouTube, and embedding those videos with their transcripts on a dedicated site page compounds the SEO value by giving models dense information rather than forcing them to infer meaning.

Resellers face a harder version of the same problem. Rishabh's priority for them is consistency, keeping titles, descriptions, and the 10 to 20 attributes behind each listing aligned across every platform a product appears on. They do gain an advantage on non-branded queries, where they can run the multi-brand comparison content a single brand typically avoids, though branded searches still favor the brand's own site. Josh described Bambi Baby's approach to content beyond the product page as "always be testing," with dedicated video and copywriting teams producing across formats at once, guided by the questions customers repeat across its brick-and-mortar locations and over Instagram.

Measuring Visibility When the Best Placement Never Gets a Click

Passionfruit measures AI search with the same funnel logic as any other channel: share of voice, citation frequency, and sentiment at the top, engagement in the middle, and organic revenue at the bottom. To connect those signals, Passionfruit built proprietary tech that ties into tools brands already use, including Google Analytics and Triple Whale, because Analytics files ChatGPT-referred visits as a generic referral rather than their own channel. One consistent pattern stands out: engagement from AI-referred visitors runs at least 30 to 40 percent higher than traditional search, since those shoppers arrive only after working through their questions in a chat. Which platform matters most varies by brand, with Claude leading for software companies, ChatGPT leading for ecommerce, and Bing increasingly showing up.

The hardest gap to close is offline. Customers regularly tell Bambi Baby staff that ChatGPT recommended the store, but they cannot measure how much weight that carried in the decision, even with a first-party pixel tracking the on-site journey and internal tools for in-store attribution that include asking customers whether an LLM sent them. His answer for leadership when traffic looks flat is to shift the KPI from raw sessions to qualified demand capture, arguing that flat sessions paired with rising conversion and order value are not a decline but the disappearance of low-intent research traffic that never converted. Rishabh frames the same case around net-new non-branded organic revenue rather than raw traffic.

Reviews, Reddit, and UGC as Trust Signals

Reputation management on Reddit comes with a warning from both speakers: do not spam it. Rishabh recommended starting with five to ten genuine engagements before scaling, ideally through real users or an agency rather than a manufactured campaign, since communities there are cracking down on AI-driven spam and AI systems already treat Reddit as a trusted source. Josh, a twenty-year Reddit user, went further, noting that users spot people trying to game the platform and that getting labeled a "shill" does more lasting damage than an unresolved bad review. His preference is to fix bad experiences at the source and let satisfied customers speak for themselves, including through influencer partnerships where real customers record long-form video reviews.

UGC on Reddit moves the needle precisely because these models treat user discussion as the highest-trust signal for whether a product actually performs. YouTube is Bambi Baby's largest UGC investment for a related reason, since transcripts give models detailed, structured information rather than requiring inference, and pairing a video with its transcript on a dedicated site page has driven several strong keyword placements. On Passionfruit's side, the firm runs managed Reddit community engagement on subreddits AI engines already cite, and Rishabh pointed to Google indexing Instagram content as a sign that content now has to work across channels, which is why he pushes brands without big budgets to make content multimodal across text, video, and images, and repurposable rather than remade for each platform.

Budget Priorities for the Rest of 2026

The clearest ROI at Bambi Baby comes from anything feeding the AI referral funnel: structured data, deep PDP content, and legitimate customer reviews, since that traffic converts roughly 50 percent higher at a higher average order value. Josh also flagged email and SMS as underrated, since owning that contact grows more valuable as discovery becomes less brand-owned, and warned against setting flows up once and never nurturing the list. On the other side of the ledger, generic top-of-funnel content and the paid spend behind it are underperforming, because AI Overviews answer those questions directly and shoppers no longer click through for a one-sentence answer, which has pushed Bambi Baby toward longer-tail, higher-intent phrasing. Through the rest of 2026, the priority order is product data and schema alongside technical SEO first, then long-form comparison and Q&A content refreshed with real updates rather than a new publish date.

Rishabh's addition is topical authority, building coverage across an entire subject rather than chasing individual keywords, which naturally captures the longer, higher-intent sub-queries that query fan-out generates. Two things he would pull back on are heavy investment in ChatGPT ads while that ad infrastructure is still immature, and tools that mass-produce hundreds of low-value articles a month, which build impressions without driving traffic or revenue and risk getting flagged as spam. Both speakers closed on the same note, that the tactics holding up across SEO, AEO, and GEO are the ones that always have legitimate content built around real customer needs, and that shortcuts producing a short-term lift eventually get recognized and penalized as models improve.