Simple. You (a company) do one thing. You do not absorb other companies and extinguish them. Somebody else does the same thing you do, but a bit different? Want to merge? Bend over, cuz here comes the part where you have to painstakingly prove why you merging with them is good, and won't subject the market to crippling hyperoptimization forces.
Walmart took control of pricing in the dairy sector causing small operations to no longer be able to compete, propped up by income from other revenue streams. Google has basically set up a value desert around itself as far as advertising goes. As a knock on, it's extinguished potential competitor/disruptor after disruptor. Facebook has absorbed and integrated personal info on a gargantuan scale with the Instagram and WhatsApp acquisitions.
Action items off the top of my head: Alphabet gets dissolved. Facebook calves off Instagram/WhatsApp. Bare minimum.
Start an investigation into how vertical integration of brick and mortar grocers and packaging Industries effects actors in that space.
Murdoch I'd have to do more research on, as he's a relatively recent arrival on my list.
Amazon gets to decide if it's going to be a logistics company, or a platform for the sellers, or it's own "brick and mortar"-like. If marketplace, no using third-party seller for order flow analysis and front running. AWS gets calved, period.
The bigger you get, the more regulators take a stake in shaping your growth, to the breaking point. Thems the breaks. Stay small if you want to do it your way. Risk becoming national infrastructure/utility as the numbers of M&A's increase.
It's not all stick though. A company can do a strategic "merge-and-release" whereby for a short period of time, a company can acquire another, help it scale, then calve off after a period. After a certain size point, this is the only arrangement by which consolidation can happen. Calving off of industrial verticals is required for further permanent merger activity to continue.
The end goal of this doctrine is to keep network effects at bay. Once you start getting snowball effect, your shareholder's needs take second place to stakeholder's needs. Period. It is insane how effective a one way wealth funnel our system can become once we stop trustbusting.
Oh, every telecom merger in the past 10 years gets audited, and gets subjected to a public review. Any found to have not significantly increased in the benefit delivered to the end user gets an asset auction, loses eligibility for bidding on municipal projects, all last mile municipal ISP bans are nulled.
I'd throw in a day in the stocks as a target of rotten vegetation in the town square for lobbyists that have been intentionally obtuse, but I really don't think that's within the realm of possibility.
All of this is liable to change if someone would be interested in hiring me for a reasonable sum to map it all out full-time complete with authority to engage in legal discovery to get what info I don't have access to that'd actually make anything in this post more than frustrated pontificating.