> ## Documentation Index
> Fetch the complete documentation index at: https://agenticadvertisingorg-feature-feedback.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# L1: Agentic advertising and the reversed data flow
> AdCP Decision-Makers module L1: what agentic advertising is, why it differs from programmatic, the reversed data flow, and how to explain it to a CMO. Reasoning only — no code.
# L1: Agentic advertising and the reversed data flow
**\~15 min** | Prerequisite: none | Free
The first thing a decision-maker needs is an accurate mental model — one you can repeat to a CMO without hand-waving. This module builds it. There is nothing to run and no agent to query; you reason through the shift and practice explaining it.
The core idea: **programmatic sends a thin request out; agentic advertising brings your data in.** A bid request carries a page URL, a device type, maybe a user ID — to a remote decision-maker that does not have the conversation context. The AI platform closest to the user *has* that context. So instead of sending requests away from the context, you bring your ingredients — brand identity, rules, goals, and (if you sell products) a product catalog — to it, and the platform generates the response.
That is why there is no creative to traffic and no segment to buy. You provide ingredients and a goal; the platform assembles the outcome.
The shift in one line — the way Ben Masse of Triton Digital put it to a room of broadcast CEOs at the egta CEO Summit:
> **From bidding to reasoning. From impressions to relationships. From milliseconds to months.**
That's the sentence to leave in a CMO's head.
Want the hands-on version? Foundations module [A1](/docs/learning/foundations/a1-agentic-advertising) teaches the same paradigm by having you query a live agent and read its response. This track assesses you on reasoning instead — same idea, no building.
## What the consumer sees
Picture a shopper who asks an AI assistant *"best 65-inch TV for a bright living room."*
* **Old world:** a banner for your TV might sit next to the chat — generic, ignorable, disconnected from the question.
* **Agentic advertising:** the assistant generates a sponsored recommendation from your **product catalog** — the list of products, prices, and images you already keep, the same feed you'd send to Google Shopping or Amazon — and your brand voice, naming the model that actually fits a bright room, in your brand's tone, with a clear disclosure that it's sponsored.
You never wrote that sentence. You supplied the ingredients — product data, brand voice, the rules — and the platform assembled the right message for that moment.
If the shopper has a follow-up (*"how does it handle glare?"*), **[Sponsored Intelligence](/docs/sponsored-intelligence/overview)** lets the conversation continue *with your brand* rather than ending at a static ad. That multi-turn, brand-to-consumer conversation — not a banner — is the deepest form of this channel, and the one worth showing your CMO.
**Agency?** This is your pitch to clients: stop trafficking finished creative, start orchestrating their data into platforms that generate it.
**Solo or SMB?** You already live this on Google Shopping — you push a product feed and the platform merchandises your products. AI surfaces work the same way: the AI writes the recommendation from your feed instead of showing a banner. No new creative to make.
## Reading list
"Why existing approaches fall short" and "from campaigns to ingredients" — the reversed data flow in plain language.
Where the two standards differ and how they work together — Sponsored Intelligence sits alongside your programmatic stack.
How a media team runs discovery, creative, execution, and reporting through one protocol.
The advertising model where the platform generates the message in the moment, for that conversation.
## Key concepts
* **The reversed data flow** — programmatic sends thin signals out to a remote bidder without context; agentic advertising brings rich data in to the platform that holds the context and generates the response
* **From campaigns to ingredients** — you provide your brand voice, rules, and goals (plus a product catalog if you're selling products); the platform assembles the ad. Better ingredients, better results
* **Not banners in AI apps** — the ad is generated from your brand data and feels native to the experience, not a display unit bolted onto a chat window
* **Additive, not rip-and-replace** — Sponsored Intelligence is a new channel that sits alongside your DSP, your agency, and your measurement tools
* **AI executes; humans stay accountable** — agents do the work, but the decisions that matter stay with people. Delegating the doing is not delegating the judgment — the cleanest answer to "but who's in control?"
## What you'll demonstrate
Sage verifies three demonstrations through conversation — the same for every learner:
1. **Explain the reversed data flow** and why bringing data to the context beats sending requests away from it. `l1_ex1_sc_reversed_data_flow`
2. **Explain it to a CMO** in plain language — "the platform generates the message from our brand data," not "banner ads in AI apps." `l1_ex1_sc_cmo_explanation`
3. **Contrast it with the model you know** — no creative to traffic, no segment to buy; ingredients and goals instead of trafficked assets and targeting. `l1_ex1_sc_no_creative_to_traffic`
## Assessment rubric
| Dimension | Weight | What Sage evaluates |
| ----------------------- | ------ | ------------------------------------------------------- |
| Paradigm understanding | 35% | Grasps the reversed data flow as the core distinction |
| Executive communication | 35% | Can translate the concept for a non-technical executive |
| Contrast with legacy | 20% | Contrasts with programmatic / IO without mismapping |
| Framing accuracy | 10% | Avoids the common misconceptions |
Passing threshold: 70%. Scores are internal — you experience the module as a conversation that continues until you have demonstrated mastery.
"I'd like to start certification module L1."