4AAVC101 - week 8

« Back to 4AAVC101

These are my notes from November 14 for 4AAVC101 at King's College London for the 2017-2018 school year. The lecturer, Nick Srnicek, is the author of two excellent books at the intersection of technology and leftist politics: Inventing the Future (with Alex Williams), and Platform Capitalism.

The usual disclaimer: all notes are my personal impressions and do not necessarily reflect the view of the lecturer.

---

The data industry

Readings

Networks of control by Wolfie Christl and Sarah Spiekermann (chapter 5)

A report from 2016 (link to PDF). A very long and thorough chapter on the data broker industry. Proposes an interesting concept: customer lifetime risk (analogous to CLV) resulting from corporations using our personal data in ways we may not anticipate (and which may harm us). Goes into the problems with hashed email addresses (they lull you into a decidedly false sense of security) which I thought was right on the money since I’ve come to similar conclusions myself while dealing with customer/partner-provided hashed emails at Macromeasures … Overall, excellent report & very uncanny to read given how much of this world I had to see in firsthand through Macro :/

The like economy by Carolin Gerlitz and Anne Helmond

On the one-way flow of data (esp likes/shares) from users (for tracking + popularity metrics) & how that’s walled off in Facebookland.

Lecture

• history of the data industry:
  • customer tracking via loyalty cards (esp frequent flyer)
    • even if we get some benefits from these cards, the companies get a lot more benefit (in terms of data on you & convincing you to keep paying them)
  • direct marketing (mail based on demographics, magazine subscriptions) - much coarser
    • acxiom got their start in this field (1969), working for Dems to send politicall messaging
  • the credit bureaus (experian, transunion, equifax), which sell to financial institutions
  • governments collecting data on citizens which is sometimes sold to corporations
• two main types of companies in this industry: the primary data sources, and the aggregators
  • primary sources:
    • consumer-facing corporations (based on purchase history, warranty forms, marketing surveys)
    • web browsing (registration data, web history, cookies, digital fingerprinting, online shopping)
    • governments, still (property, taxation, health, records)
    • cell phone service providers (mostly location), privacy regs prevent this under EU jurisdiction but common in US/Asia
    • internet of things: smart TVs, echos, cars, fitness trackers, nest etc
  • good chart from the FTC, available in this PDF report (page 2)
  • data brokers: acxiom, epsilon, equifax, experian
    • what do they sell?
    • raw data (customer emails etc) but only a small part of the industry
    • they prefer to sell service-style products on top (higher margins + they get to protect their raw data)
      • risk mitigation products (identity verification and fraud detection for businesses & individuals)
      • scoring products: credit, frailty (mortality-related actuarial services), whatever else you can imagine
      • lead generation products
• customers in the data industry
  • financial institutions for determining who’s loan-worthy (centralised data, banks pool together info)
  • employers (background checks)
  • govt (though there are more legal restrictions than with employers - they can’t collect directly, but can buy from corporations)
    • example: acxiom collaborated with US govt after 9/11
    • also the UK govt is prevented from collecting certain forms of personal data on its citizens so it has to outsource that to corporations (which is mindbogglingly inefficient imo)
  • digital ad platforms (including FB: Personicx)
    • Facebook Audience Insights will show you what they have on you and where it comes from
• dynamics of the industry: in order to make money they have to expand data collection, build increasingly byzantine products on top or erect artificial barriers to keep out competition
• problems we see in this industry
  • security breaches: acxiom, epsilon, equifax most recently
    • the worrying thing is that it’s moved from just credit card info to the stuff that’s can be used for identity theft (SSN, name, address, birthdate)
    • by aggregating so much data, they become such a juicy target for hackers, thus security breaches are basically inevitable due to structural dynamics (it’s a race that they won’t always be able to win)
  • ability to use data for discrimination in everyday life (bypassing legal protections by abdicating responsibility to the spectre of the algorithm)
    • lack of accuracy, accountability: these companies are allowed to be incredibly opaque with, say, scoring methods, very hard to challenge/correct
  • manipulation
    • companies collect data on people they know to be gullible (esp elderly) & sell that on
    • but also nudging behaviour
    • predatory pricing
• one solution: increase transparency (centralised db allowing consumers to log in & see what data these companies have on you)
  • some companies, like acxiom, are already doing this
  • but complete transparency (for tracking down original source) isn’t always possible, cus that metadata isn’t always stored
  • attention economy implications: would take us ages to, say, read privacy policies and make “rational decisions” as consumers to opt-in or not
    • also most of the time we don’t really have a choice; there isn’t an actual free market for a lot of services
    • one way to fix this: opt everyone out to begin with (like what the EU is doing with GDPR, enforceable starting may 28, 2018) but honestly that’s kind of BS, this should not be a consumer choice
  • plus transparency is a voluntary and half-hearted attempt at self-regulation in order to stave off harsher, more existential regulation
• a better solution: more strict regulations on collecting data in the first place
  • we as consumers don’t really get enough benefit from this industry to make it worth the risks
  • after all, the real point of this industry is to prop up aggregate demand by extending the reach of advertising, stretching it to its limits
• open q about long-term sustainability of these companies if digital advertising isnt growing that much?
  • think about longer-term implications for capitalism: is it dying or not? how long can we prop this shit up
• also open q about impacts of net neutrality on this industry? will ISPs just kill off the existing companies in this landscape & become the big players instead?

---

powered by jekyll. source on github.