Sure, search does really well for contextual because there is lots of intent. But most other contextual is not like that: people aren't looking to buy something.

Your normal usage isn't dog training or subaru oil changes. It's idle nonsense like whether Kim Kardashian is angry with Courtney Cox or whatever. There's not enough to sell you that's contextual. People have an idealized vision of what they spend time on. It's nothing like this productive stuff you're talking about.

> There's not enough to sell you that's contextual.

Next time you are in a waiting room flip through the magazines that are sitting there. The ads are targeted to the likely readers of those magazines.

No volume. Hard for small guy to use.

Online targeted ads are way better.

Not to be annoying, but advertising is serious money. If you think you can do good contextual advertising you will become rich very easily. Anyone will. It's hard for me to believe that no one is doing this supposedly easy and effective thing well since all the incentives are there.

> Online targeted ads are way better.

Sure, but there are a lot of ethical problems with them.

Imagine starting a service that paired individual shoppers with a passive handler to follow them around a shopping mall to build dossiers including:

    * everything they pick up and look at
    * what they eat in the food court
    * the clothing styles and sizes they try on
    * what their transportation to the mall was
    * their race, gender, age, and apparent ethnicity
    * etc...

How long would you put up with that?

Sounds like a personal shopper, which is an actual business.

It doesn't sound like that at all. It's more like the shopping center hires people to follow you around and write down all this info about you while you shop, whether you want them to or not.

Can we really be sure that current user-targetet advertising is better than contextual advertising? It seems to me that nobody wants to step out of the safe model of targeted advertising, since it does work well enough, and so there haven't been big enough attempts at contextual advertising to really say one approach is better. As an anecdotal test, I went to cnx-software.com and disabled my adblocker. The site has roughly 13 ads, 3 served by Google, the rest custom. Google tried to sell me toothpaste, while the custom ads are for things I'm actually tangentially interested in, like SoMs, embedded devices and assembly services. This kind of advertising obviously has a large overhead for the site admins right now, but I could definitely imagine an AdSense-like service that would distribute ads based on processing the site's contents.

Yes we can be sure, especially because user-targeting includes contextual signals. Google and Facebook are two of the most valuable companies on the planet because they have the science of ad targeting completely figured out.

That's like arguing heading to Alaska to prospect for gold in 1899 was a great idea because the companies selling picks and shovels in Seattle were making bank.

No. Those companies were making bank selling picks and shovels because picks and shovels are great for digging for gold, regardless of whether you actually find any.

Tired analogies and ridiculous strawman arguments aside, Facebook and Google are valuable because they have a valuable product in their advertising technology. What and how you use it is an entirely different issue.

Sometimes…sometimes not. I'm sure you've seen the blogspam articles that tell you how to fix whatever problem you're having with your computer and the first step is to download their product to do a "scan".

Sure, but what about showing you ads for WhatsApp when you’re searching for Signal?

Where does contextual advertising/influencing switch from being helpful to being gaslighting?

Aren't ads always about giving one competitor an advantage regardless of any objective benefits for the consumer? (With the exception of PSAs and the like.)