It's an unsaid general principle in business that the abilities of workers and capacities of assets are treated as the most
elastic resources when productivity is strained, and their own limits are treated as the last
hard boundary to an organization's strategy. Basically, it's assumed that the output of your assets
will be the first thing in an organization to be pushed to the limits under pressure. The assumption is that your assets can easily be pushed to produce just a little bit more when in doubt, not that they have a finite capacity and shouldn't be driven to the point of damage.
I think it's fair to say the same thing about a data analytics department's own computational resources. When you have a hammer, everything's a nail. Computation is only getting cheaper and cheaper as is the infrastructure it runs on, and the attitude of Big Data is that even if a tiny bit
more info can be gleaned for the profile of one more insignificant person who might be a potential buyer, it's worth the energy to analyze him in order to maximize profits as much as possible.
If data analytics companies find it profitable to mine bits of data we'd consider insignificant, on individuals we'd consider insignificant, and they have the resources to do so, of course a state agency would. The mentality wouldn't be "don't analyze unless their actions make it necessary," it'd be "analyze them all before
their actions make it necessary." If they already have the resources to track ten thousand people, the ten thousand first isn't a stretch, no matter how insignificant he may seem.