Your Media Plan Is Dead. AI Search Agents Ate It.
The internet your campaigns were built for doesn't exist anymore. You can't change that, but you can change how you compete with better first party data.
Here’s the uncomfortable truth: AI search agents now sit between your brand and your buyer. ChatGPT, Perplexity, Claude even Google don’t send traffic; they give answers. “Zero-click” is becoming default, and AI referrals are surging. If you’re not in the agent’s shortlist fed by structured data and credible third-party proof you’re invisible.
Publishers feel it too: when AI summaries sit on top of results, users click less. Directionally, that’s bad for anyone depending on organic discovery which is also leading to a surge or advertising come back in an unprecedented way.
Macro: ad growth is real
Global ad spend topped $1T in 2024 (first time; double-digit growth).
U.S. digital hit $259B in 2024 (~+15% YoY).
2025 is still growing (mid-single digits on a media-owner basis).
Bottom line: Budget is flowing to AI-enabled, closed-loop channels (commerce media, CTV, platforms that prove incrementality).
The three shifts you can’t ignore
1) Agents > knobs.
We’re going past “automation.” theatre AI assist; agents act planning, pacing, audience precision, creative remix continuously. Treat this as structural, not cyclical. Translation: build fast or buy fast to wire agents & specialized assistants into your stack.
2) Outcomes > proxies.
Privacy torched deterministic tracking; AI-powered probabilistic measurement made performance smarter and faster. The winning loop: first-party data → predictive models → real-time optimization → incrementally truth sets. Commerce media and CTV already operate this way; everyone else is catching up.
3) The channel map rewired.
Every walled garden wants to own “last mile” content and creative workflows (they are even good at that). Brands must own audience, experimentation, decisions, and the omni-channel loop.
Meta: Advantage+ turns budgets into a performance appliance; gen-AI shrinks production (image-to-video, brand-safe remix); “business AIs” close the loop with service.
LinkedIn (B2B): Accelerate compresses setup from days to minutes and reports sizeable CPA reductions vs. classic campaigns.
Amazon: Creative Studio + AI video/audio scale assets; AMC + DSP close the loop on and off Amazon.
Retail/Commerce media: Top-heavy (Amazon/Walmart), now going off-site with retailer data for omni-channel outcomes.
CTV: From “brand” to performance video as supply exploded and measurement matured.
Creators: Still under-weighted vs. attention; U.S. influencer spend is climbing into double-digit billions.
How AI-mediated discovery changes your odds
Assistants compress discovery → evaluation → shortlist into seconds. They privilege structured product data (specs, price, availability, FAQs) and trusted third-party signals (reviews, analysts, forums). If your data isn’t structured and your proof isn’t syndicated, you’re outside the answer set. Think SEO circa 2012 only faster, with higher stakes.
90-day playbook
Weeks 1–2: Rebase the scorecard.
Run the P&L on CAC↔LTV, incremental ROAS, new-to-brand, repeat rate. Stand up MMM + lift tests so you’re not arguing last-click.
Weeks 3–4: Make first-party data useful.
Fix consent/identity; ship server-side events to gardens/clean rooms; deploy one propensity model that actually changes bids/creative this quarter.
Weeks 5–6: Put agents on the field with guardrails.
Start with pacing, audience expansion, creative remix. Hard rules: brand style, geo/regulatory, frequency, exclusions. Weekly human veto measured against outcomes.
Weeks 7–8: Move budget to proof loops.
Retail/Commerce media on-site + off-site; hold everything to incrementality.
CTV with QR/URLs, matched-market tests, post-exposure lift.
Creators at scale with codes + modeled halo; treat them like a performance line, not swag.
Weeks 9–12: Engineer for AI inclusion.
Publish structured specs/pricing/availability; expose calculators/how-tos agents can cite; seed credible third-party validation.
Verticals punches
D2C/Retail: Treat creative like perishable inventory refresh weekly via gen-AI within brand rules; pair social scale with retail media’s closed-loop proof.
Financial Services: Clean rooms + modeled conversion make compliant performance real; own the authoritative content assistants will quote.
B2B: LinkedIn Accelerate baseline; layer ABM signals + social proof (G2/analysts).
Enterprise Retail/CPG: Treat commerce media as primary; demand off-platform activation and incrementally across CTV/social.
Hard truths towards actions
Build a composable data → decision → activation layer for first-party data that’s agentic-AI ready (real-time identity, consented event streams, feature store, privacy-safe clean rooms, server side tracking, activation into walled gardens, governance baked in).
Black boxes are fine if they deliver verify with MMM, holdouts, and independent lift (don’t outsource truth to platforms).
Speed beats perfect. Weekly creative/segment/model releases > annual debates the platforms iterate daily; match their cadence.
You don’t need perfect identity to win (Advertising UIDs is a distraction). Clean first-party signals + modeled outcomes are enough to allocate budget intelligently.
Consolidate around outcomes. If an agency partner can’t prove incrementally, they’re ornamental. Cut them.
Bottom line: This is the end of “set-and-forget” media or letting an unsophisticated media agency spray budgets. Winners run decision-first, outcome-driven systems powered by agents and fed by clean, consented data. The window is still open, but shrinking. Build it or buy it then move.
………………………………………………………………………………………………………………………………………………….
Thanks for reading. If you’re an Innovator building on AI-mediated customer/ buyer journey on either customer or vendor side or you want to upgrade your CDP/CDW/data platform into an AI agent-ready decision-activation engine, I’d love to chat.
References
Adobe, The Explosive Rise of Generative AI Referral Traffic (May & Jun 2025): AI-driven referrals up 10×+ since mid-2024; rising across verticals. Adobe Business+1
Pew Research Center (Jul 2025) & coverage: 8% CTR with AI summaries vs. 15% without; higher abandonment when summaries appear. Fortune+3Pew Research Center+3Search Engine Land+3
WARC, Global Ad Spend Outlook 2024/25: $1.08T in 2024 (+10.7% YoY). page.warc.com+1
IAB/PwC, Internet Advertising Revenue Report: FY 2024: $258.6–$259B U.S. digital (+~15% YoY). IAB+1
MAGNA, Global Advertising Forecast 2025: $979B media-owner ad revenue in 2025 (+4.9%). Marketing Dive+1
LUMA Partners, State of Digital 2025: AI, Outcomes, Media Consumption; rise of agents; commerce media & CTV as outcome engines. LUMA Partners+2LUMA Partners+2