The Real Ad Dollar Flow

How Platforms Capture Value From Brands And Creators

The Short Story Before The Long One

Every day, brands wire billions into the digital ad machine. That money doesn’t take a straight path to your screen. It winds through walled gardens (Google, Meta, TikTok), open-web ad auctions, retail media networks, and creator monetization schemes—each with their own toll booths. Understanding who keeps what is the difference between throwing budget at impressions and building a system that compounds returns.

This is a guided tour of that money trail—what brands actually spend, how the big platforms capture value, what trickles down to creators and publishers, and why transparency still lags the pace of innovation.

The Size Of The Pie

Global ad spend crossed a historic threshold in 2024 and is on track to go bigger in 2025. GroupM projects total advertising will surpass $1 trillion with growth near high single digits; eMarketer’s 2025 outlook expects about three quarters of total ad spend to be digital for the first time. In other words, most ad dollars now start and end on digital rails.

Zoom in on 2024: Statista’s aggregation suggests close to $1.1 trillion in total global ads last year and a double-digit lift in social ad spend, underscoring how creator-driven feeds are soaking up dollars. In the U.S. alone, Insider Intelligence (eMarketer) estimated $309B in digital ad spend for 2024.

Two fast-rising slices reshape the pie’s edges:

  • Retail Media—ads bought on retailer platforms using first-party shopping data—will reach ~$177B globally in 2025, overtaking TV ad revenue for the first time. Amazon leads, but China and the U.S. dominate overall growth.

  • Connected TV (CTV) has turned streaming into a performance channel with brand budgets to match; Amazon’s Prime Video is now one of the largest ad-supported streamers in the U.S., with hefty upfront commitments from advertisers.

Where Budgets Actually Go

Most marketing plans don’t publish their line items, so we rely on market share as a proxy for allocation:

  • Google + Meta still form the core “duopoly” in the U.S., collectively just under half of the digital ad market in 2023 and regaining share into 2024, with Amazon a strong third via search, display, video and retail media.

  • TikTok’s ad business keeps expanding rapidly, with global revenue growth continuing in 2024–2025 even as it experiments with more revenue-sharing products for creators and advertisers.

  • X (Twitter) remains a small line item by comparison; after steep declines post-acquisition, forecasts show a mild revenue rebound in 2025—still well below 2021 levels.

The retail media line item is the standout reallocation trend: with closed-loop measurement and shopper data, it increasingly competes with search and social for incremental dollars.

Follow The Dollar Inside Walled Gardens

When brands buy inside platform ecosystems, fees and splits are comparatively simple—and the platforms capture a large share.

YouTube (Google)

  • Long-form videos: Creators receive 55% of net ad revenue from watch-page ads under the YouTube Partner Program (YPP).

  • Shorts: Creators receive 45% of revenue allocated from the Shorts “creator pool.”

  • Scale: YouTube says it paid $70+ billion to creators, artists and media companies from 2021–2023—a rare, concrete payout figure.

Example: On $1.00 of ad revenue served against a long-form video, $0.55 goes to the creator and $0.45 stays with YouTube (after any music/licensing adjustments). Shorts flip the split.

Facebook & Instagram (Meta)

Meta uses a mix of monetization schemes (in-stream video ads, Reels ads revenue share/bonuses, performance-based payouts). The most straightforward and widely cited split is 55/45 on in-stream video ads (creator/platform), originally documented for Facebook Watch and commonly referenced across Meta’s properties. The precise terms vary by program and eligibility.

TikTok

TikTok has moved from creator funds to more revenue-sharing constructs, including TikTok Pulse, which shares 50% of ad revenue on top inventory with creators (and publishers/advertisers in Pulse placements). It also runs the Creativity Program, which pays based on performance criteria but does not disclose a single, standard revenue share. For creators, Pulse is the clearest path to an explicit split.

Twitch

Twitch standardized 55% ad revenue share for streamers under its ad revenue program (separate from channel subscription splits, which have been contentious).

The Open Web: Programmatic’s “Toll Booths”

Outside walled gardens, money flows through a supply chain—DSPs, SSPs, ad exchanges, verification vendors, data providers—before an ad reaches a publisher. Multiple landmark studies have quantified the “ad-tech tax”:

  • The ANA Programmatic Media Supply Chain study found significant take-rates across the chain and large portions of spend landing on MFA (made-for-advertising) sites, undermining effectiveness. While exact fees vary, the headline finding is that a material share of every dollar is consumed before it becomes working media.

  • The earlier ISBA/PwC study in the UK found only ~51% of advertiser money made it to publishers, with ~15% of spend previously “unattributable” in the logs analyzed. Methodologies have improved since, but the directional signal—opaque leakage—remains.

Add market structure to the mix: the U.S. Justice Department alleges Google has dominant shares at multiple layers of the ad tech stack (ad server, ad network, exchange), creating conflicts of interest that can disadvantage rivals and publishers—claims Google contests. The EU has pressed similar cases, including a 2025 adtech fine tied to self-preferencing allegations.

When The Open Web Is The Platform

Not all open-web monetization relies on header bidding complexity. With Google AdSense, publishers receive ~68% of ad revenue for content ads and ~51% for search-style units—a clear (and public) split that has served as a long-running benchmark for publishers calculating expected yield.

CPMs, CPCs, And The Price Of Attention

The price of reach and response fluctuates with macroeconomics, auction pressure, and format mix. Recent patterns:

  • Stabilization After Volatility: Skai’s Q4 2024 trends cite stabilizing ad prices across retail media, search, and social, after the pandemic-to-post-pandemic whipsaw.

  • Meta Pricing Momentum: Tinuiti’s Q4 2024 benchmarks show Meta CPMs up ~15% YoY, with Instagram spend up ~20%; Facebook spend re-accelerated into the holidays. (Tinuiti aggregates billions in managed spend; it’s not the entire market but is a robust directional gauge.)

  • Retail Media Costs: CPCs in retail media can climb quickly with demand. Walmart Sponsored Products saw a ~51% CPC increase among Tinuiti clients in Q4 2024 as more brands flooded the auctions.

  • Search Benchmarks: WordStream’s widely cited industry dataset (17k+ U.S. accounts) puts overall Google Ads CPC around $4–$5 in 2024 benchmarks, with ~7% average conversion rates—directional averages that mask big swings by vertical.

Creator-Driven Vs. Traditional Vs. Programmatic

Traditional media (linear TV, print) focuses on reach with well-established pricing and audit frameworks—but less granular targeting. Programmatic open-web offers addressability and scale but taxes budgets with intermediation and brand-safety risks. Creator-driven platforms package audience, content, and ad delivery in one place—often exchanging transparency for convenience and performance.

Two structural shifts tilt budgets:

  1. Retail Media & CTV: Closed-loop measurement (did the ad influence a purchase?) and precise audience graphs shift money from “probabilistic” to “deterministic” channels. Retail media is now a must-buy line item in many CPG P&Ls; CTV is where brand storytelling and performance converge.

  2. Social Video: Short-form and creator ecosystems have become the default top-of-funnel for younger cohorts, with plumbing (affiliate links, shops, product tags) pulling ads closer to commerce.

How Much Reaches Creators And Publishers?

Let’s simplify the math with representative, public splits:

  • YouTube long-form: 55% to creators, 45% to the platform (net of certain costs). Shorts: 45% to creators from a pooled model.

  • Facebook/Instagram in-stream video: commonly referenced 55/45 split to creators/platform; program terms vary by product (e.g., Reels, in-stream).

  • TikTok Pulse: 50/50 revenue share on premium adjacency; other creator payouts (Creativity Program) depend on views/eligibility rather than a fixed ad-rev split.

  • Twitch ads: 55% of ad revenue to streamers under its current ad revenue share construct.

  • Open Web via AdSense: publishers retain ~68% for content ads (~51% for search units).

Implication: Walled gardens now offer clearer, often more generous rev-shares to creators than what many publishers realize after the ad-tech fees of the open web—one reason why creators (and some media brands) prioritize platforms where payout math is explicit.

The Transparency Problem (And Power Dynamics)

Transparency issues show up in multiple places:

  • Programmatic log-level opacity and incentive misalignment (e.g., arbitrage and MFA supply) still siphon dollars. The ANA and ISBA studies revealed quantifiable leakage, despite improvements.

  • Platform rule changes (e.g., shifting creator bonuses, eligibility gates) can alter payouts overnight. Meta has cycled between bonus programs and revenue shares; TikTok sunsets funds in favor of rev-share products; Twitch adjusted subscription splits and ad incentives—each change reshuffles creator income.

  • Brand safety & labeling: X has faced scrutiny over unlabeled or ambiguous ad formats and content controls, spooking marketers and complicating measurement of what’s paid vs. what’s seen.

  • Antitrust and regulation: U.S. and EU actions challenge integrated ad-tech stacks and self-preferencing; outcomes may force structural separation, altering fee flows.

Net effect: Platforms with scale and first-party data have bargaining power to set terms. Creators and publishers tend to accept those terms when the audience access (and potential earnings) outweighs the loss of control.

Case Study #1: YouTube’s Two Economies—Long-Form vs. Shorts

YouTube’s monetization illustrates how format determines payouts.

  • Long-form: The classic YPP model pays 55% to creators. For channels with strong watch time, RPMs (revenue per thousand views) can be attractive, and monetization stacks with fan funding, Premium revenue, and shopping.

  • Shorts: The 45% rev-share from a pooled model makes earnings more dependent on share of total Shorts views (and music licensing tiers), improving over the old “fund” bonuses but still lower per-view than long-form.

  • Scale proof: YouTube disclosed $70B+ paid to creators/artists/media companies across 2021–2023, underlining why it remains the default “home base” for many professional creators.

Brand takeaway: Longer-form video is where creators with narrative depth monetize best, while Shorts expand reach and discovery. Allocating across both keeps creators diversified—and helps brands plan full-funnel collaborations.

Case Study #2: Cleaning Up Programmatic Spend

The ANA’s multi-year work on programmatic transparency documented fees and MFA exposure large enough to blunt performance. Advertisers who followed recommendations (supply-path optimization, MFA exclusion lists, fewer SSPs, direct deals) found higher effective reach for the same budget. Even if your exact savings differ, the direction is consistent: fewer pipes, cleaner supply, better outcomes. 

Case Study #3: Retail Media’s Closed-Loop Advantage

GroupM and eMarketer point to retail media crossing the chasm from “add-on search” to core line item, projected near $177B in 2025. Brands can target with SKU-level signals and prove impact with sales attribution—why an increasing share of search and social dollars migrate to Amazon, Walmart, and other RMNs. Amazon’s Prime Video ad product then gives retail advertisers a branding surface connected to commerce data—a powerful combo for incremental budget.

Are Creators Getting A Fair Share?

Creators have organized for better terms (e.g., the YouTubers Union/FairTube effort in Europe) and publicly pressured platforms around revenue splits and policy enforcement. Twitch’s monetization changes sparked sustained community debate; Meta’s shifts between Reels bonuses and rev-share have drawn scrutiny; TikTok’s move toward clearer revenue-sharing via Pulse was a step creators demanded. The through line: predictability matters as much as percentage.

Brands also shape fairness—by funding creators directly (sponsorships, UGC licensing, affiliate deals), not only by buying platform ads. That bypasses some platform tolls and can yield lower CPMs/CPEs at higher creative relevance—if measurement tracks incremental lift, not vanity views.

What This Means For Your Media Plan

1) Treat Walled Gardens As Different “Financial Systems.”
A $1 on YouTube long-form funds creators directly via a transparent split; a $1 in Shorts behaves differently; a $1 on TikTok may not yield rev-share unless you’re in Pulse inventory. Plan creative and KPIs by split, not just by platform.

2) In Retail Media, Expect Rising CPCs—Buy Early, Buy Smart.
Competition is tightening. Use retailer audience data for incremental reach and align to SKU distribution; diversify beyond Amazon to reduce auction pressure.

3) In Programmatic, Remove Leakage You Can Prove.
Consolidate SSPs, insist on log-level transparency, and block MFA supply. Push partners to show fee stacks end-to-end.

4) Price For Volatility, Not Just For Averages.
Benchmarks are helpful (e.g., Meta CPMs up mid-teens in Q4; Google Ads median CPCs), but your auction is situational. Build budget “shock absorbers” and automate bid/creative testing.

5) Buy With Policy Risk In Mind.
Antitrust outcomes (U.S./EU) and evolving disclosure rules (e.g., ad labeling on social platforms) can redirect fees and change inventory quality. Hedge spend across channels and keep contracts flexible.

The Money Map (A Mental Model)

  • $1 to YouTube long-form:$0.55 to the creator; ≈ $0.45 to the platform (before music/licensing).

  • $1 to YouTube Shorts: pooled, creator receives ~45% of their allocation; effective RPMs vary.

  • $1 to Facebook/IG in-stream: commonly ~55% to creator, ~45% platform (program-dependent).

  • $1 to TikTok: Pulse placements split 50/50 with eligible creators; other inventory may offer no creator rev-share (creator paid by brand directly or via separate programs).

  • $1 to Open-Web Programmatic: net to publisher varies widely; AdSense offers a transparent ~68% share for content, but multi-hop programmatic paths often see meaningful fee take rates before the publisher gets paid.

What’s Next

  • Retail Media ≈ Core, Not Edge: Expect shopper-data targeting and in-store/CTV tie-ins to pull even more budget.

  • Creator Compensation ≈ More Standardization: Platforms are converging on revenue shares and structured payouts over one-off “funds,” especially in short-form.

  • Policy & Antitrust ≈ Fee Realignment: Remedies could decouple parts of the ad stack; publishers may gain leverage, or at least better transparency.

  • Measurement ≈ Incrementality Over Clicks: As CPMs and CPCs fluctuate, brands that prove incremental lift (sales, sign-ups, brand outcomes) will win planning meetings—and budget renewals.

Bottom Line

The digital ad economy runs on trade-offs. Walled gardens trade transparency for scale and data; the open web trades control for complexity and leakage; retail media trades higher CPCs for closed-loop attribution; creator marketplaces trade standardization for creative authenticity. If you follow the money—and audit where it leaks—you can reallocate toward channels with the clearest splits, cleanest paths, and highest provable lift.