AI agents could mediate $3–5 trillion in global consumer commerce by 2030, up to $1 trillion in the US alone.¹ Agentic commerce is where AI agents shop on our behalf. It’s more than just a shopping evolution. It’s a paradigm shift that is forcing brands to go back to the drawing board.
Find out why traditional marketing funnel strategies no longer apply, and how the Agentic Commerce Grid helps brands and agencies build, execute and measure media campaigns in a way that’s relevant today.
Why the Marketing Funnel No Longer Fits Commerce
The funnel neatly assigns channels to fixed stages of the shopper journey: CTV for awareness, Google for consideration, Amazon for purchase. Budgets, measurement and teams are organized around these stages. The disconnect between teams has always been problematic, causing fragmented measurement and siloed planning; today, it’s unworkable.
The customer journey no longer follows a predictable path. Agentic commerce compresses discovery-to-purchase into a single session. This fundamentally disrupts everything marketers know to be true, and it’s a growing problem.
Traffic from AI sources to U.S. retail sites grew 393% year over year in Q1 2026, continuing the acceleration seen during the 2025 holiday season, according to Adobe.²
The neat, straight line through the traditional funnel no longer exists.
Meet the Agentic Commerce Grid
Channel, ad type, and journey stage no longer move in lockstep. We can’t continue to measure them as if they do.
The Agentic Commerce Grid is the operating system for a post-funnel world. It offers a connected view of every channel your marketing touches. Instead of asking “how much should we spend on Amazon this quarter?” brands now ask “what outcome are we aiming for, and which channel will get us there fastest?”
The Grid’s Three Axes: Channel, Ad Type, Journey Stage
The Agentic Commerce Grid maps three axes:
- Channel spans Amazon, Reddit, TikTok, OpenAI, Pinterest, The Trade Desk, and emerging platforms.
- Ad type spans sponsored products, display, video, search, shoppable, and conversational formats.
- Journey stage spans awareness, discovery, consideration, purchase, and loyalty.
Together, the three axes give you one map of where you show up, how you show up, and what job that touchpoint is doing.
Every Channel Can Drive Every Outcome
In the funnel, each channel has a distinct role. The Agentic Commerce Grid is not limited in that way.
- A TikTok video can take someone from first sight to checkout in minutes.
- A Reddit thread from an existing customer can bring in a new one.
- A ChatGPT conversation can settle a purchase decision.
That’s outcome-based marketing in practice: every channel has the potential to drive every outcome. See how leading brands are executing this across multiple channels.
Built on Your Data and Designed to Expand
The Agentic Commerce Grid is built from your own data, shaped around your channel mix, and pointed at your business goals. Learn how to structure your omnicommerce strategy with the New Omnicommerce Blueprint. The Grid is never finished: as new platforms and formats emerge, they slot into the Grid instead of forcing you to build a new measurement system.
The Agentic Commerce Grid vs. The Funnel
Budget by Outcome Instead of by Channel
As channel roles flex, budgets should do the same. Instead of, “how much should we spend on Amazon this quarter, and which ad types?” The Grid asks a different question: “What outcome are we aiming for, and which channel, or mix of channels, will get us there fastest?”
Outcomes Over Reach
In retail media measurement, funnel-led reporting gives you platform-specific metrics such as reach or impressions. This tells you who saw the ad but not what it sold. The Grid ties verifiable commerce outcomes (growth, margin, share) directly to the exact channel and touchpoint that drove them.
One Workflow and a Single Source of Truth
Using multiple platform logins limits your control over cross-channel performance. With the funnel, each team sees only its own channel’s performance. The Grid replaces that with a single workflow and one source of truth. Optimization happens in real time, and everyone sees the whole picture.
“Marketers managed channels; the next era is managing outcomes. Brands need an intelligence layer connecting every touchpoint to commerce.” — Melissa Burdick, Co-Founder & President, Pacvue.
How Pacvue Prism Powers the Agentic Commerce Grid
Pacvue Prism is the commerce media platform that connects all your marketing channels to the outcomes that matter for your brand.
First, it automatically pulls all your channel data into one place, so you don’t have to manually export reports from Amazon, TikTok, and everywhere else, and stitchthem together yourself.
Second, it runs one AI system across every channel, looking at them through one lens. It can tell you that spend on Reddit is driving purchases on Amazon three days later, something you’d never see if each channel had its own separate reporting tool. This also means you can optimize from one screen, not by hopping between a dozen dashboards.
Third, it plugs into the data you already have from your customer data platform, sales and margin numbers, inventory, and digital shelf health. This is how it ties ad performance to actual business results.
Three Pillars Powering the Grid
Prism organizes all of this across three pillars, making cross-channel commerce truly measurable for the first time:
The Reach Engine allows you to scale awareness and discovery across audiences through programmatic platforms like Amazon DSP and The Trade Desk. It expands reach beyond owned retail media into the full programmatic ecosystem.
The Commerce Growth Lever connects social and discovery platforms such as TikTok, Reddit and Pinterest directly to commerce outcomes like sales, margin or new-to-brand. These platforms compress the discovery-to-purchase phase into single instances. Prism shows how social engagement results in measurable revenue.
The Differentiation Layer captures emerging behaviors in frontier formats such as OpenAI, ShopsenseAI, Firework, and tvScientific by Pinterest. Conversational, shoppable, and streaming formats reach customers often when intent is highest.
Every channel reports against the same outcomes, so it becomes one input to the shared picture. Prism unifies retail media, search, programmatic, social, CTV, and frontier formats under one AI agent, strengthened by access to your data lakes, CDPs, CRMs, and custom AI agents via MCP Data Centralization.
The Numbers From Prism’s Early Access Participants
Pacvue’s early access participants (25+ Alpha brands) to Prism have seen:
- -23% CPM
- +12% total sales
- +24% new-to-brand growth
Results are from brands who participated in Pacvue Prism’s early access program. Individual results will vary based on factors specific to each user’s environment. The performance improvements referenced are not guaranteed and do not represent typical or average results for all users. Pacvue makes no representation that any particular user will achieve the same or similar outcomes.
Going beyond the numbers, Alpha users describe the change in their overall approach:
“Most organizations assign budgets at the channel level, but in the long term, this could eliminate that process,” says Nic Jones, Director of Media at PetIQ. “You’re no longer dealing with channel-level budgets. Now you’re dealing with what is the goal of these dollars. That in itself is really powerful.”
Ready to Power Your Brand’s Agentic Commerce Grid?
Uncertain whether your brand is AI-ready? Start with the Agentic Commerce Readiness Scorecard. It’s a diagnostic tool to assess your current maturity and identify where to focus first.
Ready to see Prism in action? Book a demo with our team.
Footnotes
¹ McKinsey. “The Agentic Commerce Opportunity: How AI Agents Are Ushering in a New Era for Consumers and Merchants.” https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-agentic-commerce-opportunity-how-ai-agents-are-ushering-in-a-new-era-for-consumers-and-merchants
² Adobe. “AI traffic grows but retail sites lag in AI search visibility.” https://business.adobe.com/blog/ai-traffic-surge-retail-sites-not-machine-readable
