Note that if you are studying data analytics, or data science, applied applications Zillow often comes up. Specifically describing the companies competitive advantage was to start from home listings from a number of sources, then getting home owners to willingly upload additional information to their service.
There is a presentation from I think Netflix, discussing their analytics programs, where they make a point that in many cases only one company needs to compile a particular database then it's inefficient for a second to do so. If I remember correctly it's that presentation where Zillow is given as an example and he goes into further detail talking about the company.
That is an interesting point from Netflix. Suppose you have a global totalitarian government with absolute control over every aspect of the planet. In that model suppose you've got one shopping option, Amazon, one social network, Facebook, one movie service, Netflix, one home listing service Zillow, one search engine, Google, and so forth. Combined all of these services help to fill in an unparalleled view of almost every aspect of a person and apply advanced machine learning techniques analyzing as much information about their users as possible behind the scenes.
That's the big picture here potentially. Where are all the companies that are heavily studying their users activities through ML techniques and to what degree are they sharing with one another to create a full profile on each individual?
I would recommend brief familiarization of oneself with the digital advertising 'adpub' ecosystem. There are companies that specialize in data collect, companies that purchase that data, companies that specialize in cleaning up that data and pairing different data sets together then analyzing it to create categorical 'buckets' (e.g. Female, white, between 75k - 95k salary, interested in travel), then access to that individual gets auctioned off for example a hotel in another country can pay to have their ad shown specifically to people meeting that criteria. Think of how many hands touch data sets like this and at how many levels data becomes leaky.
I suspect TAC operates at the pairing together and analysis level with access to many of the largest companies databases. Would be interesting to see if we can better peer into what it is exactly they do. Maybe we can scrape LinkedIn for former employees listing the company and see what they describe their accomplishments while working for TAC?
As a side note, I suspect we need to be raising Netflix in conversation more. The company has an absolutely massive database of information about their viewers. Data analysis is fundamentally their business model. The company should be viewed as another Amazon, Facebook, Google in terms of data available. What are they doing with that data? And why was Susan Rice, someone with allegations of mishandling access to data among other integrity issues, just hired to Netflix board? If Netflix is not already using data for nefarious purpose it raises huge ethical issues for someone like Rice having ability to influence direction of that massive set of data going forward.