Today, I spoke at the W3C improving web advertising business group about the economics of digital ad identity. Thread🧵 on some takeaways on the future of digital ads. 1/
Cross-site identity (via cookies) allows for behavioral targeting. Importantly, identity allows for less sexy but *critical* functions like ad frequency capping, ad effectiveness measurement & attribution—all at scale. 2/
What is the value of cookies? In the status quo, most studies and data agree that cookies create value: ads get 50-70% less revenue without cookies. 3/
What does advertising look like without cookies? Ad spending will fall by *less* than 50-70%, because the market will adjust. But, the ad revenue pie will shrink, and will be divided more unequally. 4/
Analogy: What happened to my industry (post-secondary education) when we offered a worse product (Zoom/hybrid)? Our industry's revenue pie shrunk, Harvard is fine, & the lower-end is haemorrhaging. 5/
Post-cookies, digital ads have four tentpoles. Value of identity advantages the walled gardens & sign-in (i.e. hashed email identifier solutions). Context will struggle to compete: especially general & news content as well as small sites. 6/
How can privacy & modern advertising coexist? Good news: Advertisers care about audiences not individuals. Bad news: Privacy implies some hard tradeoffs. 7/
What are "the birds"? Various proposals (notably by Google Chrome’s Privacy Sandbox, Criteo) try to preserve the value of ads for the open web while protecting user privacy. This is the difficult, but important work of this W3C group. 8/8
• • •
Missing some Tweet in this thread? You can try to
force a refresh
“Abandon all hope, ye who enter here” - Dante’s Inferno👿
Online display ad experiments are hell. They are also a proving ground for field experimenters, & have much to teach us. The guide is organized into the nine 9 circles of 🔥hell🔥 as applied to #displayad#fieldexperiments
🔥Circle 1🔥 Display ad effects are so small🤏 that observational methods fail🤦♀️. Ad effects explain so little variation in ad outcomes, that they get swamped🌊 by unobserved confounds. Like Dante entering the inferno👿, we resign ourselves to the necessity of experiments.😭😭
The @guardian featured an opinion piece about how the #GDPR is failing to protect privacy.
The piece serves as an unintentional object lesson of the same. THREAD 1/ theguardian.com/commentisfree/…
Here is the excerpt where the author decries prevailing opt-out practices alongside the Guardian’s consent menu doing the same.
Note: The @ICOnews states that this menu is not #GDPR compliant (“NO" should be as easy as "YES"). 2/
When I VPN as a French user The Guardian interacts with 42 third party domains (listed below) and loads 122 third party cookies.
Note: All this arises without my opt-in consent. 3/
Google just released a more details about the study showing *publisher revenue* falls 52% without a cookie. They run an experiment disabling cookies for a random sample of users on Google Ad Manager's top 500 publishers in terms of programmatic revenue. 1/ services.google.com/fh/files/misc/…
Now, we see that results vary by publisher (below). While the average publisher’s revenue falls by 52%, the median is 64%. Very few publishers see a loss less than 20% while a large number experience losses exceeding 70% (!). 2/
Despite precise aggregate estimates (95% CI <1%), the individual site estimates are statistically imprecise for smaller sites. Focusing on the top 200 sites reveals less dispersion. Now, only a 1 of 200 sites loses less than 10% of revenue. 3/