full-spectrum customer analytics: sketches

gal and i have been thinking a lot about data and privacy.

below are some ramblings for/about potential projects.

  • a facebook algorithm that recommends new friends for you based on common traits among your existing friends. it analyzes your photos for tattoos, pets, siblings, colors, and other identifiers.

the problem with this, and with some of our other data project ramblings, is that they’re not different enough from what actually happens, on the internet and off. we do develop mental models that let us take shortcuts to finding friends—shortcuts based on arbitrary things that only sometimes accurately reflect someone’s capacity to be a good friend. just because someone has lots of cat photos doesn’t necessarily mean that they’re kind ∴ good friend or a total slob ∴ inconsiderate ∴ bad friend, even though we might have developed those shortcuts in our brains. facebook algorithms probably incorporate some of the same shortcuts when it recommends friends to us. all of which is to say: still trying to find the interesting pressure points in this scenario.

  • a performance piece in which we work for a data company and pitch our analytics services to a client. our company also has a data derivatives branch that repackages information we get from our clients so we can sell it to our other clients. data bundles are graded from top-shelf AAA down to B3.

“Our services help clients optimize product distribution, target prospective data points more effectively, grow revenues, and harvest more data.

The keystone of our business is our in-depth “Customer Portrait” parsed from analytics bundles we have access to through our Insights Initiative—a partnership between Deloit, Complast, and Faceboop.

Through our Data Derivatives branch, we also work with clients to develop their own data-backed revenue streams so they can capitalize on their customer-provided insights. We have expertise structuring data packages that maximize profits and preserve privacy through our proprietary Randomized Identifier Reassignment algorithms.”

we actually started exploring this project in more depth. for the classroom setting, we wanted to create “customer portraits” of some of our classmates, which involved using facebook and google to find information about them, their interests, their families, and where they live. i got increasingly uncomfortable with the kind of work we were doing to put together the portraits, even though we knew they’d ultimately be used in a performance critiquing the exact thing we were doing.

a conversation about this scenario comes up in every art theory class i’ve ever taken—the one about institutional critique and when it becomes not critique anymore, when it just reinforces the thing it’s supposedly critiquing. usually, i’m talking about it from a spectator’s perspective. it felt really different to actually have to make a decision about whether to continue a project. similar icky feelings stopped me from pursuing elements of my “white people in coffee shops” project, and they also stopped me from pursuing a weird internship.

i’m still trying to figure out what to do with this information. do i need a system for deciding when to stop and when to keep going? what do i gain by stopping? what do i lose by stopping?

additional reading/watching:

Leave a Reply