I have no idea. The story about tailoring ads to individual consumers that the tracking companies sell is catnip to marketing departments, but I don't have high confidence that the advertising industry is better at the statistics needed to evaluate what they are doing than, say, psychologists were before the replication crisis.

While this works for Google Search (to my knowledge, search ads are still exclusively contextual) most contextual ads only work in limited situations - random blog websites or things like quizlet don't really have any related products you could advertise on based on the domain, while people might still be included to (for example) purchase an Amazon item they viewed a few days ago.

Advertising is zero sum in some idealized economists’ fantasy world where everyone knows everything about every product.

In real life, targeted ads add value by matching products and consumers that would not have found it in the jungle of information.

The competition among advertisers is zero sum.

If A has one ad and B has another had and there's a single consumer, A and B spending more on ads doesn't change the outcome. It's zero-sum, no?

Or to make it even more obvious, imagine all advertisements in existence are on a giant billboard. You could imagine people allocating the space perfectly in some fashion. However some people would spend more money and plaster their ad on top of someone else's. This would result in others doing the same. More and more money is being spent but it's zero-sum.

But what if C comes up with a new product that is perfect for you? You’ll never know, and C will go out of business. A and B will continue to dominate.

Now with mass marketing, the biggest budget has the most reach, so it’s hard for C to break into the market.

With targeted ads, C at least has a chance to reach the customers who might love it.

Targeted ads gets us one step closer to a more dynamic economy that serves every individual perfectly and one step further from entrenched corporates selling the same old middle of the road crap to everybody. Crap that nobody loves or hates.

Even in your scenario it’s still zero sum with respect to advertising dollars spent and advertisements delivered.

No, it’s not zero sum. There’s now a better solution where customers are more optimally served by a new product than before. Let’s call the increase in value of this new solution X.

Money was spent on ads to make it happen. Call it Y.

If X > Y, value was added. If not, then new company C eats the difference and loses money and/or goes bankrupt.

No, I’m talking about the advertising dollars. You’re taking about something else. Ultimately the amount of attention at any given moment is fixed yet the amount of money that be spent pursuing such attention is unlimited.

Just because money being spent has a positive ROI doesn’t mean advertisements are not zero sum.

By your logic ads can be a disruptive experience and hence negative sum. In fact I’d be surprised if isn’t the case. Most people find ads to be annoying.

If ads provided value to the customer, you wouldn't have to pay extra for ad-free services -- you'd have to pay extra to get the ads.

There's a pretty strong argument that it doesn't, at least in many cases. Cory Doctorow's How to Destroy Surveillance Capitalism is one long-form answer to that question.

https://onezero.medium.com/how-to-destroy-surveillance-capit...

I'm not convinced he's entirely correct. Marginal improvement over other methods may be useful, and there are some cases, notably political advertising aimed at dissuading target groups of voters from participation, which may be at least sufficiently effective to be useful.