>> Let's assume the worst of Google and conclude they abused their search monopoly. Innovation in search was stifled. Competitors were damaged and achieved only a fraction of their potential. For instance, Yelp might be worth $15B instead of $3B in a world without "evil" Google.
Abuse manifests itself in many ways, only one of which is the stifling of innovation.
Let us consider the lookalike audience feature in Facebook and Google. You, the engineer, are tasked with improving the efficiency of the system. You are told that a 1% improvement in the effectiveness of this targeting produces like a gazillion dollars in the company's bank account.
You start by running some numbers to understand the effectiveness of the current system. You need some metrics to do this, obviously. One metric is the click through rate of an ad when it targets a certain lookalike segment, which is probably easy to measure. But then you see if it lead to a purchase - but how can you do this? In the world of ad-ROI measurement, this is like the last mile - the so called click attribution problem.
You come up with the following suggestions to your bosses:
1. Capture more of the data going through the inter-pipes. Create Google Analytics so you know who goes to which website. Have people logged in to Facebook permanently, often without their knowledge. Make apps which ask for all kinds of permissions which they don't need. Create devices which are capable of tracking anything you want, and you control the switch remotely.
2. Buy every company in sight which might have collected some reasonable amount of data
3. Start making data triangulations. Collect and purchase datasets which allow more of this.
4. To get people to spend more with you, expand the size of the smallest sized lookalike segments (larger surface area for ad targeting). This will at some point take you back to steps 1, 2 and 3.
Imagine all this going on at Google and Facebook scale.
But wouldn't this all happen even when there are no monopolies? Perhaps, but there would be a lot of potential for data brokerages. And those data brokerages, not unlike markets of goods and services, will start assigning accurate prices for personal information. The monopolies are preventing these data brokerages from being created. They are cutting out the middlemen, so to speak. And more often than not, they are doing it with the sort of impunity that was inconceivable even a few years ago. Right now, no one really knows if their personal information is worth anything. But if it wasn't worth anything, then Facebook to a larger extent and Google to a smaller extent, wouldn't be nearly as profitable. There are many startups being created today which don't want to call themselves that, but are glorified data brokerages with the explicit goal of being acquired by a tech giant to get a big payout in return for the consolidation of often private information.
For example, around 2012 I heard of the idea of 'privacy lockers' where people will put their personal information and all companies will have access to it, but for a price. Imagine that! If that had actually come into existence, Facebook would have been acquired by the privacy locker company of course. Which in some ways only leads us back to the original problem.
So the next question is whether this consolidation in and of itself is a bad thing? I don't know, is a nuclear weapon dangerous when it is controlled by a madman? It is worse in some ways. We can't be certain about either the potency of the weapon (private data) or the sanity of the people in charge (corporate motivations).
There will be a reset button at some point. Splitting up the companies into smaller entities seems like a hasty conclusion today. When (not if) these tech giants cross some unacceptable privacy boundaries, the resulting backlash is going to result in a lot of wistful "wish we could go back in time when splitting was still an option" thoughts.