Figure 2 — How fast each channel is getting more expensive (in real terms)

Channel Ranking

Each bar shows the quality-adjusted price compound annual growth rate (QAP CAGR) for a channel — how much more an advertiser pays per year for the same unit of useful attention, after subtracting bots, scams, and the audience that can't afford your product. Counterintuitively, the smallest bars are the healthiest channels: direct mail, out-of-home (OOH), and host-read podcasts are barely moving (~1-6%/yr). The biggest bars are the channels where advertisers are losing the most ground every year. TikTok is the steepest at +37%/yr, which means a dollar spent on TikTok ads in 2025 buys roughly 1/5 the useful attention it bought in 2020. Click any bar to see why.

← Healthier (price stable)Degrading faster →
Channel
0%25%50%/yr
QAP CAGR
Structural Resilience
OOH
+3%/yr
Direct Mail
+1%/yr
Podcasts HR
+6%/yr
Newsletters
+6%/yr
Volume-Build Masking Decay
Retail Media
+6%/yr
CTV
+8%/yr
Podcasts DAI
+1%/yr
YouTube
+8%/yr
Mature Decay
Search
+24%/yr
Meta
+23%/yr
LinkedIn
+23%/yr
Accelerated Collapse
Linear TV
+17%/yr
Display
+6%/yr

Dot position = quality-adjusted price CAGR. Dot size = 2025 impression volume (logarithmic). Channels grouped by archetype.

Reference Card

ChannelTypeQAP CAGREffective CPM (2025)
OOHA+3%/yr$7
Direct MailA+1%/yr$457
Podcasts HRA+6%/yr$35
NewslettersA+6%/yr$37
Retail MediaB+6%/yr$2
CTVB+8%/yr$90
Podcasts DAIB+1%/yr$23
YouTubeB+8%/yr$16
SearchC+24%/yr$519
MetaC+23%/yr$82
LinkedInC+23%/yr$83
Linear TVD+17%/yr$111
DisplayD+6%/yr$20

Quality-Adjusted Price (QAP)

The real cost of advertising per unit of useful attention. Calculated as nominal CPM (cost per thousand impressions) ÷ (Attention × Density × Legit Share). Two channels can have the same headline CPM but radically different real costs if one is full of bots and a downmarket audience that can't afford your product. The chart shows how fast that real cost has risen each year from 2015 to 2025.

Concrete example

Say you spent $1,000 on TikTok ads in 2020 and got 100 units of useful attention from real humans who could actually buy your product. In 2025, the same $1,000 only buys you about 14 units. The other 86 evaporated into bots (~35% invalid traffic), inattentive scrolling (3.5s active attention per ad vs 13.5s on premium video), and an audience that skews dramatically downmarket (only 12% of TikTok users earn over $100K).

Why smaller bars are better

A small bar means each year you get roughly the same value per dollar as the year before — the substrate is stable. A big bar means the substrate is degrading and the auction is squeezing more dollars out of the same eyeballs. Most channels are in positive territory because everything is degrading at least a little. OOH at +3%/yr is healthy. Meta at +23%/yr is broken.

The TikTok paradox

TikTok's nominal CPMs have actually been falling — Tinuiti measured a 30% drop in Q1 2025 — because advertisers are bailing faster than the auction can recover. But the quality is collapsing even faster than the price is falling. You're getting "cheap" ads on a substrate that's deteriorating faster than the discount. TikTok skipped the healthy era most channels had — it scaled directly into decay.

Bar length vs bar color

Bar length measures how fast the channel is degrading. Bar color shows the structural archetype: green = substrate can't be manufactured, yellow = volume-build masking decay, orange = mature decay, red = accelerated collapse. The two dimensions overlap but tell different stories.

Click any bar

The detail panel shows the four underlying quality factors at 2025: attention coefficient, audience density, legit share, and the channel's projected nonviability crossing dates.

Primary Sources

Meta 10-K filings (2015-2025)Alphabet 10-K filings (2019-2024)Nelson-Field, 'Cost of Dull Media' (2025)Adelaide Attention UnitsPew 'Americans Social Media Use 2025'FTC Consumer Sentinel 2024 Data Book