# Webinar: Agentic Commerce - your next 90 days Source: https://www.youtube.com/watch?v=DTG5hADdkBs Recap page: https://rapidrecap.app/video/DTG5hADdkBs Generated: 2026-02-20T12:01:41.612+00:00 --- ## Quick Overview The presentation outlines a strategic 90-day framework for tackling Agentic Commerce, emphasizing that the current behavioral shift is real but the playbook is unproven, requiring brands to focus on building understanding through disciplined portfolio experimentation rather than concentrated optimization or overinvesting in website replatforming. **Key Points:** - 30-45% of US consumers use AI for purchase research, indicating a significant behavioral shift that is already happening, with AI referrals showing 4-16x higher conversion than organic search. - The primary challenge in Agentic Commerce is ownership, as AI relies on data and input from SEO, Product Data, PR & Comms, and Partnerships, with the recommendation layer being the most commercially valuable part. - The recommended 90-day framework involves four workstreams sequenced by evidence quality and effort: 1) Understand (map current AI position), 2) Measure (build trustworthy attribution), 3) Experiment (run disciplined tests), and 4) Build (develop compounding capabilities). - Brands should avoid immediately chasing proprietary AI shopping agents or creating dedicated AI SEO teams, as the platform landscape is fragmented and unstable, and the playbook is not yet proven. - Key SEO signals like domain authority, content quality, authoritative backlinks, and structured data remain relevant, but rankings are increasingly tied to AI mentions, and citation selection/rate data is different from traditional organic metrics. - Experimentation should focus on testing hypotheses across third-party content quality, structured product data, authoritative coverage, and platform-specific testing. - The ultimate goal is to move beyond mere AI mentions to achieving AI recommendations, which is where the highest commercial value concentrates. ![Screenshot at 00:00: Title slide for the webinar "Agentic Commerce: What's Real, What's Premature, and What You Should Do," setting the stage for a discussion on AI's impact on commerce.](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-00-00.jpg) **Context:** This presentation, titled "Agentic Commerce: What's Real, What's Premature, and What You Should Do," delivered by Amir Ouki from BOI (Board of Innovation), addresses the rapidly evolving landscape where AI, particularly generative AI like ChatGPT and Gemini, is influencing commercial decisions. The speaker emphasizes that while the shift in consumer behavior is significant (with 30-45% of US consumers using AI for research), traditional analytics methods often fail to capture this influence, leading to an undercounting of AI's true impact by 2-3x. ## Detailed Analysis The webinar on Agentic Commerce establishes that a real behavioral shift is occurring, with 30-45% of US consumers using AI for purchase research, resulting in AI referrals having a 4-16x higher conversion advantage over organic search, despite only accounting for less than 1% of total web traffic. The core challenge identified is ownership, as AI recommendation outcomes depend on inputs from SEO, Product Data, PR & Comms, and Partnerships, with the intersection being the most commercially valuable area. The speaker cautions against overinvesting in website replatforming or building proprietary AI shopping agents prematurely, as the playbook is unproven and the landscape is unstable. Instead, a disciplined 90-day framework is proposed: 1) Understand the current AI position via structured query audits and third-party presence audits; 2) Measure attribution by segmenting AI traffic in GA4 and tracking branded search volume; 3) Experiment with disciplined tests focusing on third-party content quality, structured product data, authoritative coverage, and platform testing; and 4) Build compounding internal capabilities like AI representation literacy. The presentation stresses that traditional SEO signals remain partially relevant (domain authority, content quality), but citation tracking must be viewed differently from commercial metrics, as AI models are trained on vast, often non-commercial data. ### Introduction to Agentic Commerce - Welcoming attendees from global locations like Oxford, Lisbon, Toronto, Paris, and Washington; defining Agentic Commerce as top-of-mind for everyone. ### What 'agentic commerce' actually means - Three core components with different time horizons—Conversational Recommendation (live at scale), Agentic Task Execution (easily accelerating), and Embedded Commercial APIs (infrastructure phase). ### The Evidence Base - What the data actually shows: Confidently, 30-45% of US consumers use AI for research; AI referral traffic is <1% of total web traffic but has a 4-16x conversion rate advantage over organic search; Global market projection is $3-5T. ### Brand Presence vs. Recommendation - Positioning brands on a spectrum from Mentioned (AI knows you exist) to Described (AI explains your product) to Recommended (AI recommends you), noting most measurement tracks the former, not the latter, where commercial value concentrates. ### What to ignore (for now) - Don't overinvest in website replatforming for AI readability; Don't abandon SEO (Google still sends 3-40x more traffic); Don't treat AI citation tracking as commercial metrics; Don't implement universal checklists; Don't concentrate investment on a single AI platform. ### 90-days - Understand your current position (Phase 01): Run structured query audits across ChatGPT, Gemini, Perplexity; Document AI responses (mentioned vs. recommended); Audit third-party presence (Wikipedia, review sites, comparison sites); Identify gaps; Produce a baseline. ### 90-days - Build measurement you can trust (Phase 02): Set up AI traffic segmentation in GA4 by tagging referrals; Weekly performance review of AI-referred sessions/conversions; Track branded search volume; Track AI Mode impressions in Google Search Console; Separate AI conversion reporting from organic. ### 90-days - A portfolio of disciplined experiments (Phase 03): Experiment A (Third-party content quality); Experiment B (Structured product data); Experiment C (Authoritative coverage); Experiment D (Platform-specific testing). Experiment rules include one hypothesis per experiment, pre-defined measurement criteria, a 90-day minimum observation window, and documenting what doesn't work (avoiding conclusions based on a single platform). The goal is to learn fastest, not act first. ![Screenshot at 00:00: Title slide for the webinar "Agentic Commerce: What's Real, What's Premature, and What You Should Do," setting the stage for a discussion on AI's impact on commerce.](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-00-00.jpg) ![Screenshot at 02:06: Slide comparing the old purchase funnel \(Awareness to Loyalty\) collapsing into a 'Single AI Conversation' that handles research, comparison, evaluation, recommendation, and purchase.](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-02-06.jpg) ![Screenshot at 03:30: Slide detailing the three tiers of agentic commerce: Conversational Recommendation \(live at scale\), Agentic Task Execution \(easily accelerating\), and Embedded Commercial APIs \(infrastructure phase\).](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-03-30.jpg) ![Screenshot at 06:00: Slide titled 'The evidence base: what the data actually shows,' comparing what can be said with confidence \(e.g., 30-45% of US consumers use AI for research\) versus directional/uncertain projections \($300B-500B projected US agentic commerce by 2030\).](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-06-00.jpg) ![Screenshot at 08:23: Slide illustrating the spectrum of AI interaction: Mentioned \(AI knows you exist\), Described \(AI explains your product\), and Recommended \(AI recommends you\), noting commercial value concentrates on the right side.](https://ss.rapidrecap.app/screens/DTG5hADdkBs/00-08-23.jpg)