You'd be surprised how a lot of marketing budget is wasted.
Especially in brand advertising, where there is no directly measurable ROI.
Toothpaste, paper towels, cars, cereal - they spend gazillions on those brands.
But think of this: most 'ad spend' is spent on completely commoditized products - I'll guarantee you the 'no name' toothpaste is as good as 'Crest' or whatever.
Our economy is made up of huge 'soft monopolies'.
The grocery stores, pharmacy chains, 'big brands' like P&G - they are all systematically slow, barriers to entry are surprisingly high (you can make toothpaste easily, but you cannot build a brand easily). Car companies, entertainment channels. Banks.
Talk to some of the front line marketing people buying ads. They are often clueless. They have no idea often if the money is well spent or not. But the have a 'budget' and 'have to spend' it across various channels. They know they need to 'go social' so they buy ads from 'known brands' like Facebook and Google.
Now - the other source of more profitable ads are simply things where you can reasonably get conversions online. Anything for your household that you can buy at Ebay, Amazon etc., i.e. consumer goods.
And then the scammy stuff.
Some of the above complaints about the value of ads have merit - that said - they may not be making products that are amenable to online ads.
While it's very hard to measure the effect of ads trying to create this loyalty, it doesn't mean that it's not worth it.
I've discovered that for myself where I saw ads on Facebook that I didn't click, but later saw a product in store and tried it out. If you just follow conversions, you'd say that the Facebook ad was worthless, whereas it actually influenced my buying decision.
I'd argue that most money on ads is well spent, but to be effective, you must have a huge budget. People need to see your ad multiple times, so that they'll be familiar with a product. If you don't have that budget, it'll be very hard to use ads effectively. But that's no fault of FB or Google, that's just psychology.