It's not fractions of a percent. It's massive increases which is why FB and Google are two of the most valuable companies on the planet.

They have 1st-party access to user profiles which is why they're so good compared to independent parties which use more anonymous profiles. This is why 3rd-party audience data is poor and will get poorer while Google and Facebook will be more entrenched because of GDPR and similar laws. So there will be a move to context for some ads, but not most of the ones you see.

> It's not fractions of a percent. It's massive increases

Do you know a number on the difference between targeted and untargeted search (not display) ads? If I search for "instant pot," I'm probably looking to buy an instant pot. How much does knowing my entire search and browsing history help the advertiser? I'm sure Google has these numbers, and equally sure they aren't sharing.

It looks like there may be big differences for display ads (2-3x?), but for search?

> It looks like there may be big differences for display ads (2-3x?), but for search?

Keep in mind that display ads are wildly ineffective to begin with. Targeted ads may raise your click through rate from 0.2% to 0.6% but that isn't good enough to justify the massive privacy loss and use of bandwidth, server time, engineering talent, etc...

> engineering talent

That's the saddest part. A lot of talented people could be working on important problems, but instead they're trying to make people click on ads a tenth of a percent more, so they will buy Crest slightly more often than Colgate.

I guess it's no worse than going into finance: just another way for someone with talent and ambition to make a boatload of money while providing almost no value to society (and sometimes negative value). Still, one could hope that talented people would sacrifice a bit of lucre to do some good. The inventors of Arpanet don't have rocket companies, but I bet they're living comfortably.

No value to society? Every single business uses marketing to grow, including the one that employs you if you have a job. How about the shareholders who benefited? That includes pensions, retirements, insurance and other funds that you have an interest in.

And what about all the other stuff that you get for free, including the majority of the content on the internet? What about all the other services like Gmail and the technical projects that have been opensourced? What about all the people who got rich and went on to invest in or start their own companies?

There's no "right" thing to work on. It's like asking why do we have astronauts in space [1] when there are problems here on earth. The truth is that it takes all kinds of people to make this world work and we're all better off if people get to do what they want to do, and hopefully it happens to be what they like doing too.

1. http://www.lettersofnote.com/2012/08/why-explore-space.html

I referenced "Crest vs. Colgate" to suggest that quite a bit of marketing is a zero-sum game. If people drink more Pepsi, they'll probably drink less Coke. If they eat at McDonalds more, they'll probably eat at Burger King less. How much money do these companies waste shifting consumers between equivalent products?

I'm not against ad-supported businesses, but I'll bet Google would have made plenty of money off plain old search ads, probably enough to make Gmail free for everyone. If not, they could have charged a few bucks a month, or shown untargeted ads. In any case, they wouldn't have had an enormous database of PII, just waiting to be hacked or mined in more malicious ways when the ad bubble pops (e.g. credit reports, insurance).

> There's no "right" thing to work on.

Really[1]? We're paying hucksters 5-10x what we're paying teachers now. If that's what the market demands, then the market is twisted.

[1] http://www.metrolyrics.com/wernher-von-braun-lyrics-tom-lehr...

Anecdotally, my mom worked for P&G when I was growing up and she said the main difference between the brands was which one was on sale, sold more than the other. In a way the price was the biggest reason for selling toothpaste, not marketing.

Yes really. There's no rank of value. Go work on what you want to work on. Changing the value of teachers is a good goal and there are plenty of edtech companies being funded so if you care about it then go fix it.

I prefer that we all have the freedom that you can actually choose to do that, or not.

> There's no rank of value.

Wow... Thats cormac mccarthy level bleak but at least honest and I hope not to live in the world your labor may bring into being.

EdTech is mostly just a scam to siphon public school money to private companies. I'd rather just lobby my school district to pay teachers more.

It was an example, but you seem to be suggesting that some people have a more worthy cause than others?

What's the most "right" thing then and why aren't you working on it already? I assume curing cancer is somewhere on that list? Let us all know when you'll be getting to it.

Well, I certainly wouldn't put adtech in the same league of significance as space exploration.

And plenty of people wouldn't find any significance in space exploration, or whatever your career is. Subjective judgements are pointless overall. All that matters is that we have the freedom to choose.

> And what about all the other stuff that you get for free

It's only free if you think that the details of your life aren't worth anything (I specifically object to advertising targeted to individuals, I have no problem with ads similar to ones you find in a magazine or print newspaper).

Without any discussion, ad companies asserted that they have the right to spy on users and track them in almost every area of their lives online and off. I think we are finally starting to push back in the other direction. Europe has GDPR and I think it's only a matter of time before we have something similar in the US. IMHO companies should only be allowed to collect and use information on individuals in ways that each individual decides to allow.

> There's no "right" thing to work on.

But there are plenty of wrong things to work on.

Clickthrough rates don't matter and have been outdated for a decade. Econometric models showing return on ad spend is what makes the difference.

Yes things are different with search ads. What you're describing is called "intent" and there's nothing more clear than what you query, although it can and does get boosted by history. I didn't think this discussion was limited to just search though.

That's kind of what I thought: Google didn't have to turn to the dark side until it bought DoubleClick. Sure, they wouldn't have become one of the biggest companies in the world, but they could still have made plenty of money. If they had taken that course, I'd still be using their products.