blog: Don Marti


don't preempt me bro

21 April 2019

Some surveillance marketing organizations have suggested adopting a Federal privacy law in order to preempt the California Consumer Privacy Act. Preemption would be bad if it actually happened, but the fact that they're trying for it is the best endorsement I have ever seen for the California Consumer Privacy Act. If I wasn't a CCPA fan before, I am now.

In my humble opinion, preemption is the wrong direction. Privacy regulation should be complicated enough to impose significant transaction costs on database marketing practices. State-level privacy regulations are a start, but what about county or city ordinances? User tracking allowed on alternate sides of the street on different days of the week?

Why would I want to see costs and complexity imposed on the surveillance marketers? I'm going to leave the political stuff out for now. From a selfish point of view, as an individual considering buying stuff, I am going to get ads, and I'm going to get them matched to me in three ways.

  1. Context. Placed on a resource I'm interested in using, like a magazine article or a bus bench.

  2. Search. Matched to search results when I look for a product or a service, like a Yellow Pages ad or a Google search or Maps ad.

  3. Personalized targeting. Matched to me based on something the advertiser knows about me.

On the Internet, many ads are placed using a mix of these techniques, and it's hard to split out how a real-world marketing budget is allocated across them. And information originally collected based on context can leak and start getting used for personalization. But the technical and regulatory environment affects how much money advertisers choose to invest in each one.

As the recipient, or potential customer, the three ad placement methods affect me in different ways. Ad money allocated to context is a subsidy for something I want to use, whether it's local news coverage or an ad-supported public restroom.

Ad money allocated to search is almost as good. I'll use a search engine more if it gives helpful results, so search advertising also pays for something I want.

Personalized targeting, though, is a problem. Instead of paying to support something I want, the advertiser is paying to reach me as an individual. The fact that my information is in somebody's database is a risk to me, but a source of revenue for the database owner. It's a classic Negative Externalities problem. Besides, anything spent on this stuff does not go to pay for the ad-supported resources and search services I really want.

Ad-supported cultural works have positive externalities, when they're re-purposed for other uses. The "Star Trek" advertisers got their money's worth in 1966-1969, but people are still watching the show today. Kurt Vonnegut quit his job as a car dealership manager because he sold stories to Collier's magazine.

As a member of the audience for advertising, I win when I can help move the marginal advertising dollar from personalized targeting to either context or search, because a fraction of the money that gets moved pays for something I want, some of it is likely to create positive externalities, and none of it gets spent on creating risks for me. Regulation is a piece of the solution, and a mess of confusing regulations could be more effective in raising the relative price of personalized targeting than a single set would.

People's intuitions about marketing practices are economically sophisticated.

  • People often choose to pay attention to ads that carry economic signal.

  • People are quick to develop banner blindness and other habits to avoid low-signal advertising.

  • People choose not to invest a lot of time in low-effectiveness ways to protect their personal information, but pick up on measures seen as effective, such as Do Not Call.

People who grow up in ad-heavy economies learn the economics of advertising like people who grew up playing ball learn physics.

What we need to see from privacy regulation is

  • increase the transaction costs of negative-externality advertising practices.

  • credible promise of reducing risks, to atract mass participation.

Privacy regulation has to have the confusion and cost from the advertiser side increased, in order to balance out the risks and costs imposed on the audience side, and shift ad budgets.

Four Steps Facebook Should Take to Counter Police Sock Puppets New law of headlines: any piece with Facebook should in the headline is not worth reading. Come on, EFF, the police put undercover officers into schools and workplaces all the time. Why should a mass-market social platform be any different? This is just the high school weed dealer's if you ask them if they're a cop they have to say yes.

Silicon Valley-Funded Privacy Think Tanks Fight in D.C. to Unravel State-Level Consumer Privacy Protections Employees of companies that back this: Is facepalm-induced neck trauma a work-related injury?

Craig Silverman at Buzzfeed continues to write the absolute best #contentMarketing for #progressiveWebApplications: Popular Google Play store apps are abusing permissions and committing ad fraud

Carole Cadwalladr blasts tech titans at TED: Your technology is “a crime scene” (It's not "their technology" though. It's thoroughly owned by criminals and domestic terrorists. High-profile "tech" execs just pay the power bill, and they're only "titans" of resting and vesting.)

With Epsilon deal, Publicis bets on first-party data for survival With Epsilon news, Publicis makes themselves the first to-do item on my CCPA opt-out list.

European Commission’s Giovanni Buttarelli on state of GDPR adoption: ‘Even ticking a box does not necessarily mean consent is freely given’ I have tried to figure out how much you need to know about web ads to give informed consent. I don't know enough myself, but I'm still learning.

Adobe announces deeper data sharing partnership with Microsoft around accounts Sweet, another one for that CCPA to-do list.

Daphne: Moderating Facebook At Barely Minimum Wage
then: hardware is difficult, manly work, and software is a straightforward office task we can hire low-paid women for.
now: software is difficult, manly work, and content moderation is a straightforward office task we can hire low-paid women for.

Almost half the Bay Area’s residents want to move, survey shows Those of you who want to stay: what's your secret?