Databases Reference
In-Depth Information
was thinking within the architecture of the browser window: LSOs are, as the name
suggests, local cookies stored outside of the browser, in my case at rraley/Library/
Preferences/Macromedia/Flash Player/#SharedObjects, and thus not deletable from
a browser toolbar.
2
In basic terms, LSOs are tracking devices within the Flash player
that override the user's security preferences and are set without her knowledge and
consent. There are applications such as Flush and BetterPrivacy that will ostensibly
manage and clean out LSOs, but their most pernicious aspect is their capacity to
“respawn” tracking cookies with data stored in Flash; that is, Flash Local Storage is used
to back up the HTML cookies for the explicit purpose of restoring them within seconds
after they are deleted. These zombie cookies—and this is certainly their effect, as
manual deletion and even the aforementioned tools are essentially futile—are made
possible by what Adobe Systems insists is a “misuse” of Local Storage, though it is worth
noting that the privacy settings panel on Adobe's site is notoriously difficult to read,
appearing as a demo rather than as an actual window.
3
They have not then been invisible
to me alone, though the larger issue of data collection continues to receive more public
attention in the wake of investigative reports such as the
Wall Street Journal's
“ What They
Know ” series.
4
The immediate purpose of LSOs, along with traditional and third-party cookies, is
online behavioral advertising, the economic potential of which is no doubt clear: con-
sider the speculative value of the uniquely numbered cookie assigned to each machine,
one that collates ostensibly nonpersonal behavioral information in order to produce a
closely approximate demographic portrait including age, gender, location, educational
level, income, consumption habits (purchasing and reading), sexual preference, and
health issues.
5
The “audience management experts” of Demdex, Inc., for example,
transform the profile of a common user into one of a unique individual by combining
the ID code from a single machine, one that holds a summary record of browsing and
search history, with offline data including census information, real estate records, and
car registration.
6
As John Battelle puts it, this information is producing “a massive click-
stream database of desires, needs, wants, and preferences that can be discovered, sub-
poenaed, archived, tracked, and exploited for all sorts of ends.”
7
Online behavioral
advertising produces a dynamic, flexible, and perfectly customized audience, consti-
tuted by the microtargeting of the intents and interests of consumers on a massive scale.
In practical terms, if a consumer happens upon but fails to make a purchase from a
particular retail site that aligns with her profile, that microtargeting can become retar-
geting, which means that ads for an item she has viewed will be pushed to other non-
