In the first season of Mad Men, Harry Crane does something nobody at Sterling Cooper takes seriously. He asks to run a television department. The other partners see TV as a novelty, a distraction from the real business of print advertising. They give him the department mostly to keep him quiet. By the end of the series, Harry is the most powerful person in the building.

The IAB/PwC Internet Advertising Revenue Report tells almost the same story. When the first edition was published in 1996, digital advertising barely existed. It was tiny, almost a joke. Thirty years later, the 2025 report shows U.S. digital ad revenue hit $296B, up 14%, and that was without the Olympics, World Cup, or an election to boost spending.

The department nobody wanted is now the entire business. And the fastest-growing part of it is video, up 11% to $82B.

Let's break it down into 6 big questions:
1) How big is the digital ad market now?
2) How big is video advertising?
3) Who is doing the actual ad buying?
4) Which companies are winning the video ad market?
5) How is the way we buy advertising changing?
6) What’s next?

How big is the digital ad market now?

Digital ad spend (IAB/PwC):
1) 2005 - $13B
2) 2015 - $60B  (↑ 375%)
3) 2025 - $296B  (↑ 395%)

Right time, right place: Betting on digital in 2009 was prescient, but at least there was a $23B market estimate. When we bet on streaming in 2016, nobody was even counting yet.

Share by format:
1) Search - 39%
2) Display - 28%
3) Video - 27%
4) Other - 4%
5) Audio - 3%

YoY growth by ad type:
1) Video - ↑ 25%
2) Search - ↑ 11%
3) Audio - ↑ 10%
4) Display - ↑ 10%

How big is video advertising?

Digital video ad spend:
1) 2017 - $12B
2) 2018 - $16B (↑ 36%)
3) 2019 - $22B (↑ 34%)
4) 2020 - $26B (↑ 21%)
5) 2021 - $40B (↑ 49%)
6) 2022 - $47B (↑ 21%)
7) 2023 - $54B (↑ 14%)
8) 2024 - $64B (↑ 18%)
9) 2025 - $74B (↑ 15%)
10) 2026P - $82B (↑ 11%)

Share by video type:
1) Social video - $32B (39%)
2) Streaming TV - $29B (36%)
3) Online video - $21B (25%)

Growth rate for 2025-26P:
1) Social video - ↑ 20%
2) Online video - ↑ 17%
3) All video advertising - ↑ 17%
4) Streaming TV - ↑ 15%

I find it hard to believe that online video (OLV) is growing faster than streaming. In our annual market sizing, I used several numbers from eMarketer, and they have streaming growing at a slightly higher rate.

Either way, social video is a rocket ship.

Who is doing the actual ad buying?

Share of ad spend (Advertiser Perceptions):
1) Independent agency - 41%
2) Brand direct - 30%
3) Big 5 agency HoldCo - 28%

Note: Omnicom acquired Interpublic Group in 2025, reducing the number of Big Agency HoldCos above to five.
1) Dentsu
2) Havas
3) Omnicom
4) Publicis Groupe
5) WPP

Ad spend change (2019-24):
1) Brand direct - ↑ 21%
2) Independent agency - ↓ 4%
3) Big 5 agency HoldCo - ↓ 16%

The brand-direct number is the one to watch. In five years, brands went from controlling roughly a dime of every ad dollar to nearly a third. The same platforms that are winning the video market (YouTube, Meta, Amazon) are also making it easy to buy without an agency middleman.

Which companies are winning the video ad market?

Digital video ad revenue (Kagan):
1) YouTube - $21B
2) Meta - $14B
3) TikTok - $14B
4) Amazon - $6B
5) Meta - $14B

YouTube generates more video ad revenue than the next two competitors combined. The top three (YouTube, Meta, TikTok) are all social/user-generated platforms. The traditional media companies that actually produce premium content are fighting over what's left.

How is the way we buy advertising changing?

Two ways.  First, upfronts are growing slower than the rest of the ad market. More money is flowing into last-minute TV buys and automated digital buying instead.

Projected upfront growth for 2024-25 (Media Dynamics):
1) Streaming - ↑ 18%
2) Broadcast - ↓ 3%
3) Cable - ↓ 5%
4) Total - ↑ 5%

Second, more and more digital ads are being bought by software, not people.

Programmatic share of digital advertising (excluding search):
1) 2021 - 89%
2) 2022 - 87%
3) 2023 - 84%
4) 2024 - 87%
5) 2025 - 90%

CAGR (2021-25):
1) Programmatic - ↑ 13%
2) Non-Programmatic - ↑ 11%

What’s next?

Three forces will shape the ad market over the next 3-5 years.

1) AI becomes the buy side. Programmatic is 90% of non-search digital. The next step is AI that buys the ads, makes the creative, and measures the results, all without a human touching it.

2) Consolidation accelerates. YouTube will do $21B in video ad revenue alone, more than the next two combined. Brand-direct buying tripled to 30%. The holdcos are merging (Omnicom/IPG) or shrinking (WPP, Dentsu). Fewer players, bigger share.

3) A bridge between streaming and social video. Everyone talks about streaming vs. linear. The real convergence is streaming + social video. Together they're $61B and growing 17%. But most brands still plan them separately, with different teams and different budgets. The companies that figure out how to optimize across both screens will have a massive advantage. As linear TV fades, it will become clear that the winning formula is streaming TV + social video.

What to watch: Whether AI grows the pie or concentrates it, and whether the economy holds long enough for the transition to play out.

Harry Crane asked for a department nobody wanted and ended up running the future. Thirty years ago, digital advertising was that department. Today, at $296B, it is the business. The question now is whether video, AI, and the convergence of streaming and social will create the next Harry Crane moment, or whether the big players who already own the screen will be the only ones left standing.