I would challenge this. It's is a technique in the sense that manufacturing cars is Toyota's technique. We would still call them an auto manufacturer and refer to this as a market, at least casually.
Most social media (and also netflix, spotify... so maybe just "media) is feed driven. The anchor to these feeds is data gathering. The data gathering edge (along with network effects) is what makes a $100bn worth $100bn. It's what makes the product work. A twitter feed, e-commerce specials page or targeted advertising widget that is customised well is an entirely different product to an untargeted generic.
If Tesla pulls off self driving, the difference between Tesla (worth $400b on expectation alone) and any other manufacturer their size (probably >$10bn) will be data.
Data may not be the only thing that will be responsible for self driving, but it is proving to be the limiting factor in many of the most commercially important cases.
The OP is all about how revenue of marketing companies shouldn't be larger than the revenue of final goods companies. But there is clearly no such restriction on self driving cars, it's a final product all by itself.
If you mix both into a "data gathering market", you have no means at all to analyze it.
Data is constantly growing new uses, which can go from economically marginal to central very quickly. Data gathering just isn't something that is a distinct market all on its own. The same dataset could feasibly power an ad widget, recommendation playlist or whatever.
I'm not sure it's even possible to discover if profits are concentrated there, the market can look exactly the same either way.