Built for retail media. Ready for agentic commerce
AI-driven commerce is reshaping how shoppers discover and buy. Moloco Commerce Media already powers your retail media business with catalog depth, real-time decisioning, and measurement. We’re building on that foundation to support emerging agent-led shopping experiences.
Moloco’s customers include:
Agentic commerce opens two frontiers
AI referrals to retail sites grew 4X year over year in 2025. As more shopping journeys begin on AI platforms like ChatGPT and Gemini and finish on retailers’ sites, retailers have an opportunity to earn recommendations from agents and build trust with customers. {[data-attribute="init-tooltip"][tooltip-content="<p>Source:<br><br>EMARKETER, “AI is growing quickly as a retail referral channel. Here’s how retailers can win.” Q1 2026.</p>"]}
Where consideration sets form
Before shoppers reach your site, AI agents use structured product information to decide which retailers to recommend. The goal is to make your products easier to discover while protecting what sets your business apart.
Where habits are earned
Shoppers form habits around experiences that feel timely, relevant, and helpful. Contextual intelligence can help retailers deliver those experiences and build stronger, direct relationships with shoppers.
Reasoning engines recommend what they can verify
When a shopper asks an agent for a standing desk that fits a small apartment, the engine isn't matching keywords — it's reasoning over structured attributes: dimensions, compatibility, use-case fit, availability. Sparse or stale catalog data no longer just lowers conversion. It risks exclusion from consideration entirely.
Make your product data agent-ready
Retailers integrating with Moloco already structure and enrich their catalog through our feed, including custom fields for the details that matter for each item. That work pays off twice:
- Today, Moloco’s AI models use it to power personalization, relevance, and product search on your site.
- In the future, AI shopping agents will be able to use the same enriched product data to understand, compare, and recommend your products.
Use one catalog foundation across channels
The catalog work already powering your retail media business does double duty:
- SEO helps products rank in search; structured product data helps reasoning engines evaluate them for recommendation.
- The same structured attributes that power onsite personalization also help external agents understand your products.
Control how signals are shared
Moloco’s approach to signal governance:
- Freely share product facts, including specifications, pricing, and availability.
- Selectively provide curated outcome insights without exposing the underlying data.
- Keep proprietary behavioral data, return rates, and the learning engine protected.
Contextual AI meets shoppers in the search-and-browse path
Shoppers have spent 20 years building search-and-browse habits — a chat bubble in the corner doesn't undo that. Chatbots wait to be discovered and require explicit action to find.
Onsite Intelligence
We believe the better model is ambient intelligence — assistance that appears the moment behavior signals uncertainty, and disappears the moment confidence returns.
Our vision: Moloco's real-time decisioning reads those in-session signals and surfaces contextually relevant product intelligence at the exact moment of friction.
Triggered by behavior, not buttons
Onsite interactions generate behavioral signals that offsite agents cannot access, including:
- Time spent reviewing product specifications
- Variant toggling
- Repeated movement between product pages
Grounded in data only you hold
Frontier models keep improving at general reasoning, but they don’t see which of your SKUs works for which use case. That knowledge lives in:
- Returns data
- Support patterns
- Post-purchase behavior
Built to earn the direct habit
When shoppers choose a store directly, that relationship can:
- Help protect margin
- Generate the proprietary data that compounds your advantage
What's live today — and what we're building next
Here's what MCM can do right now — and the capabilities currently in development.
Live in Moloco Commerce Media today
Catalog normalization and product understanding
Real-time behavioral decisioning for onsite personalization, retrieval, and ad relevance
Demand and creative automation workflows
Capabilities in active development, subject to change
Structured product data that helps offsite agents discover and understand products
Tools for controlling how signals are shared and used
Monetization based on outcomes from agent-led shopping experiences
Three priorities for retailers building now
Catalog depth is what compounds
Frontier models can't learn which SKUs work for which use case. That knowledge lives in your data — and it's what gives reasoning engines the confidence to recommend you.
Build for onsite habits and offsite rankings at once
The catalog work that powers onsite intelligence is the same work that makes you legible to external agents. One investment, two frontiers.
Learning velocity is the advantage that compounds
By the time agent traffic is meaningful, retailers who started early will have run dozens of experiment cycles. That knowledge can't be bought or compressed.
Resources
The Dual Frontier: A Retailer's Framework for Agentic Commerce
A practical guide to building onsite AI advantage and offsite discoverability, with a 15-minute readiness diagnostic.
Frequently Asked Questions
What is agentic commerce, and why does it matter now?
Agentic commerce refers to shopping journeys shaped by AI platforms such as ChatGPT, Gemini, and Perplexity rather than by traditional search or direct visits. It matters now because these platforms are already shaping which retailers get recommended and trusted—and those patterns may become harder to change over time.
How is ambient intelligence different from a chatbot?
Chatbots wait to be discovered and require explicit action to find them. Ambient intelligence — as we envision it — triggers on behavioral signals of uncertainty and surfaces contextually relevant help at exactly that moment. The interface may still be conversational; what changes is when and why it appears.
Do we need to build new infrastructure to get started?
Moloco Commerce Media can work with your existing catalog and first-party data, although some catalog updates or technical integration may be required. Moloco Commerce Media stands up your dedicated instance of commerce media AI models, campaign management, and ad-serving infrastructure. The commerce media AI models in your dedicated instance are trained exclusively on your data. This same foundation powers your retail media business today and can support agentic commerce in the future.
What should we share with agent platforms — and what should we protect?
Our view is that a three-tier approach can help determine what to share and what to protect. Product specifications, pricing, and availability can be shared freely; curated insights from outcome data can be shared selectively to provide useful conclusions without exposing the underlying data; and raw behavioral streams, attribute-level return rates, and the learning engine itself remain protected.
How will we know whether agent surfaces drive real results?
Multi-touch attribution models can undercount the contribution of newer surfaces like agent-assisted sessions. Our holdout-based incrementality framework uses control groups to better isolate net-new revenue from agent-assisted sessions.
