The 20% Toll Booth on Streaming Ads

While Google and Meta guard their ecosystems, The Trade Desk thrives by playing referee. Its model shows how enabling both sides of a market can build loyalty without needing exclusivity. With that said, marketplaces are HARD.....builder beware.

Author

Nick Horob
September 18, 2025

Today, I'm looking at The Trade Desk (TTD) They offer an independent software platform where advertisers and agencies buy digital ads across the open internet, earning revenue by taking a percentage fee on ad spend.

A few things that stand out to me about them:

  • They make money by taking a cut (~20%) of ad dollars that flow through their platform.

  • They’ve got marketplace-style benefits: more buyers attract more publishers, more spend generates more data, and better data keeps advertisers coming back.

  • At the same time, they don’t own distribution. They don’t control YouTube or Hulu inventory, so they rely on integrations.

  • That neutrality is part of their pitch to advertisers: “We’re not selling you our media, we’re helping you buy the best media.”

A dash of reality:
It’s hard to build both sides of a platform like this. You need to simultaneously convince buyers, sellers, and partners to show up. The Trade Desk pulled it off by:

  • Positioning as the neutral “enablement” layer rather than a media owner

  • Building industry standards (like Unified ID 2.0) that others had reason to adopt

  • Turning customer education into a growth engine

The Trade Desk is one of the most impressive businesses that I've covered. With that said, as I mentioned above, building a marketplace-like business is extremely hard. They are so powerful that many people, myself included, get drawn to the business model. But I now fully realize the difficulty of the model so I'll let others innovate in the space!

With that, I'll see you tomorrow.

Nick

PS - If you've ever used The Trade Desk as an advertiser, I'd love to chat with you as connected TV advertising is something I should know more about than I do. Shoot me a note!

TL;DR

  • Independent ad-buying platform with strong Connected TV presence, earning fees as % of spend (~20% take rate).

  • Strength: scale, neutrality vs walled gardens, sticky retention, AI-powered decisioning.

  • Weakness: agencies can multi-home, no exclusive inventory, subject to privacy/walled garden risks.

  • Lesson: Build enablement platforms, lean into standards, make education part of growth.

The 30,000-Foot View

  • Business model: Demand-side platform (DSP) for agencies/brands, earns platform fees on spend.

  • Key stats: ~$2.68B TTM revenue, ~79% gross margin, ~$417M net income, ~3,500 employees, market cap ~$22B.

Company History

  • 2009: Founded by Jeff Green & Dave Pickles

  • 2016: IPO at $18/share

  • 2021: Launches Solimar (first-party data)

  • 2022: Launches OpenPath (direct supply)

  • 2023: Launches Kokai (AI-first UX)

  • 2024: Launches Ventura OS concept, reports ~$12B spend

  • 2025: Joins S&P 500, new CFO

Show Me the Money

Highlights:

  • High software-like margins (~80%)

  • Expanding operating leverage (10% → 18%)

  • Net cash balance sheet

  • Capex rising for infra/CTV initiatives

Financial Data

Metric

2022

2023

2024

TTM

Revenue

$1.58B

$1.95B

$2.45B

$2.68B

Gross Profit

$1.30B

$1.58B

$1.97B

$2.13B

Gross Margin

82.2%

81.2%

80.7%

79.4%

Ops Profit

$0.11B

$0.20B

$0.43B

$0.48B

Ops Margin

7.2%

10.3%

17.5%

17.7%

CapEx

$0.08B

$0.05B

$0.10B

$0.17B

Net Debt

($1.45B)

($1.38B)

($1.92B)

($1.69B)

The N.O.O.B. Nine — Competitive Powers

The Nerd Out on Business Nine is made up of Hamliton Helmer's famous "7 Powers" of competitive advantage (Scale Economies, Network Economies, Counter-Positioning, Switching Costs, Branding, Cornered Resource, and Process Power) combined with two of my own (Data Flywheel and Distribution Advantage).

Power

Score

Rationale

Branding

4/5

Strong B2B thought leader

Data Flywheel

4.5/5

More spend → better AI → more spend

Process Power

3.5/5

Consistent shipping & ops leverage

Scale Economies

4/5

Spread infra costs, better bidding

Switching Costs

3/5

Agencies can multi-home

Cornered Resource

2.5/5

No exclusives, but UID2 adds value

Network Economies

4/5

Buyers & supply attract each other

Counter-Positioning

4/5

Neutral vs walled gardens

Distribution Advantage

4/5

Deep agency ties, MSAs

Average Score: 3.7/5 - Durable moat, nearly across the board high marks

Memorable Marketing

  • UID2 (2020–): Privacy-first identity standard → positioned TTD at center.

  • OpenPath (2022): Direct supply path → transparency win.

  • Kokai (2023): AI-first branding → improved stickiness.

  • Ventura OS (2024): Smart TV OS concept → sparked ecosystem buzz.

Takeaways: Build standards, productize education, use partners as distribution, name major releases, tease future concepts early.

AI Uses & Opportunities

  • Current: Bidding, forecasting, measurement (Koa, Kokai).

  • Next: Creative scoring, predictive frequency caps, retail media auto-optimization, supply path AI, AI-assisted onboarding.

Bumps in the Road

  • Privacy regulation & cookie deprecation

  • Walled gardens’ privileged data & supply

  • Scrutiny of DSP take rates in CTV

  • Execution risk from scaling complexity

Your Swipe File

  • Build transparent enablement platforms

  • Turn education into growth engine

  • Use standards to influence markets

  • Stay vigilant on take rates & complexity

  • Don’t rely on exclusivity, win on outcomes

  • Don't underestimate marketplace difficulty