Media monetization vs. internet advertising

Structural problems

The internet is structured in favor of ad networks. Ad spend grows approximately at the rate of inflation, but the inventory of pages on which those ads can appear grows with each new Instagram post (about 100MM per day).

Internet advertising is far more automated than print, but the benefit goes to intermediaries and buyers. On average, publishers receive only about half of what advertisers pay for the advertising that appears in their publications. Exchange fees and platform fees (both supply side SSP and demand side DSP) take the bulk of the difference.

Newspaper ad revenue fell off a cliff in the 2000s, matched by a corresponding rise in Google and Facebook revenue. (Start the thread at the top.)1

Regulatory problems

Regulatory pressure is reducing ad networks’ and advertisers’ ability to surveil and target users through cookies and other tracking mechanisms.

California Consumer Privacy Act and Nevada SB 220 require disclosure of tracking, and the option to disable tracking or the sale of tracking data.

GDPR cookie policy is even more stringent, especially following a 2019 enforcement change requiring informed opt-in before any tracking cookies can be set. Following that change in enforcement, publishers saw just 10% of visitors opt-in to tracking.

An upcoming California ballot measure may further strengthen the CCPA with GDPR-like rules requiring informed, op-in consent for tracking.

Legislation inspired by the CCPA is under consideration in a growing number of states.

Technical problems

Browser vendors are responding to consumer pressure and removing cookie tracking features ad networks and advertisers depend on to surveil users across the web. 2

Firefox and Safari have been blocking third-party cookies and other tracking on which ad tracking networks depend for some time.

A 2019 study reported that “about 73% of the ads shown on a Safari browser do not have a cookie associated.” And, “users who are browsing using a mobile device (either mobile phone or tablet) are less likely to have a cookie associated than users browsing using a desktop or laptop.”

Chrome, the last major browser that still supports third-party cookies will soon block those cookies as well.

Unrelated to cookies, it’s also worth considering the growing use of advertising block lists that substitute for editorial judgement and community accountability. Those block lists are said to suppress coverage of social justice and women’s health content by blocking ads on that content.

Some solutions

Some in the ad industry are working to find alternative means of “end-to-end addressability” despite tightening legal regulations and technical limits.

Australia is set to require Google and Facebook to pay publishers for the news shared on their networks. In the US, some are suggesting regulating Google’s ad market (spearheaded by the same lawyer who argued Facebook is a monopoly, though the two have been the target of criticism for some time), and the effort is getting some official traction.

Meanwhile, NYT, WaPo, and Vox are all developing their own first-party advertising solutions. NYT is offering “45 new proprietary first-party audience segments to target ads.” The data comes from tracking users across NYT properties (no third-party tracking), correlation of geoIP to zip code demographic data, and progressive profiling efforts.

Dutch broadcaster NPO has dropped tracking cookies entirely. Digital ad revenue is up 60-80% YoY overall, and one of their smaller properties is up by 92% YoY. This due to both better results with contextual-based targeting over behavioral, and reduced costs by eliminating intermediaries. “Because the contextual ad server doesn’t rely on tracking, it renders the thicket of middlemen largely obsolete; the money goes straight from advertiser to publisher.”

Proof point for ads without tracking

Advertising in podcasts totaled up to an estimated $479.1MM in 2018, according to the Internet Advertising Bureau. This is notable because podcast advertising cannot be tracked using the mechanisms advertisers and ad networks use elsewhere on the web.

The same IAP report estimates the market will grow to exceed $1B in 2021. Spotify’s $100MM acquisition of a highly ranked podcast this year buoys those numbers, and at least one person thinks it was a low-ball deal.

More interesting, a full 62.5% of podcast advertising in 2018 was sold on a quarterly or annual basis. Cost per thousand subscribers and flat fee pricing models dominate the space, while cost per acquisition is too small to meet the reporting threshold. Unsurprisingly, brand awareness and branded content campaigns are growing fast.

The podcast ad market is still a fraction of the $14B newspapers grossed in ad revenue (print and digital) in 2018 and the $49B they brought in for 2006. Still, the willingness of advertisers to move forward without tracking, to do long-term sponsorship deals, and to focus on branding over acquisition/conversions is an interesting sign for where the market might go in the future.

I’m not so foolish as to argue we roll the clock back on the internet, but I am interested in how what we value changes over time. I see the growth of the commercial podcast market as a sign that advertisers are again valuing the trust and accountability that media organizations have with their communities.


  1. Craigslist has long been a target of blame here as well. There’s something to that, but it’s complicated. And all media, not just those that ran classifieds have seen downward revenue pressure due to changes in advertising structure brought on by the internet. ↩︎

  2. The elimination of third-party cookie support won’t stop first party services (Facebook, for example) from tracking and targeting you while you use their services on their site, but CCPA and GDPR provide protections about how that data can be used and give users a right to delete it. ↩︎