4 ms·
I've been collecting examples of where the ads that I see are based on extremely simple algorithms of the type that could have easily been supported 30 years ag
by lkrubner 7y ago
I've been collecting examples of where the ads that I see are based on extremely simple algorithms of the type that could have easily been supported 30 years ago, and yet I keep reading articles that suggest that the advertising industry is deploying sophisticated tools to target ads to me. I wrote about this recently:
-------------------------
Despite much talk about Machine Learning and AI improving advertising results, what I’m seeing is getting worse and worse. Despite billions invested, the ads shown to me are much less relevant than that ads that I saw on the Web 10 years ago.
I hired 3 developers from Fullstack Academy. They were all great, so I went and checked out the website, curious about the curriculum. And now, every website I go to, I see an advertisement for Fullstack Academy. (See screenshot.)
I’ve been writing software for 20 years. I’ve written semi-famous essays about software development. I am not going back to school. I do not need to go to a dev bootcamp. So why show me ads, as if I’m thinking of going to school?
For the last several years I’ve been seeing articles about the surveillance economy. In theory, advertisers know more about me than ever before. In theory, they know about my entire life. And yet, the ads I see are less targeted than what I used to see online 10 years ago.
http://www.smashcompany.com/business/when-will-machine-learning-and-ai-improve-advertisings-ability-to-target-people http://www.smashcompany.com/business/when-will-machine-learn...
- dijksterhuis 7y agoI willing to take a punt and guess that the models they’re using are using short term/isolated data, not 10 years worth of your entire browsing history.
- plaidfuji 7y agoHaha, yeah. When n=1, both human and computer will say “keep doing what you’re doing!”
- foldingmoney 7y agoApparently you didn't get the targeted ad telling you that advertisers' dark patterns have become so sophisticated that free will is literally fiction now. /s, obviously
- onion2k 7y agoSo why show me ads... To keep the brand in your head so you post about it on Hackernews.
- lkrubner 7y agoThat’s a “just so” story. You’re looking at something that is easily explained by incompetence, stupidity and irrationally, yet you’re working to transform it in your head into something rational. Take a moment to think of the money they’ve wasted and what do they gain? How likely is a sale? Did Fullstack imagine this scenario when they authorized their marketing firm to spend this money? Or is the marketing firm simply trying to spend money so they can bill for something?
- onion2k 7y agoI was being a bit facetious about you posting on here, but the point I was making was serious. In advertising there's a thing called the "effective frequency"[1] which is the number of times you need to see an ad before it has an impact on you. Obviously this series of adverts has worked on you - you know the brand and you use it as an example of which ads you remember. If the company is advertising in order to raise the level of engagement they're getting that's a fail; if their ads are intended to get people talking about the company that's actually a pretty good result. There are more reasons to advertise your business that simply "getting more sales". Indirect communication is very useful. [1] https://en.wikipedia.org/wiki/Effective_frequency https://en.wikipedia.org/wiki/Effective_frequency
- NeedMoreTea 7y agoMatches my experience. I'm reminded of 80s and 90s sales lead phone lists - those used to be marketed as precision means to reach your choice of age, job-level, city, income etc. I once worked for a company that tried a few of these, from allegedly fresh, first generation data. They were universally crap, with the same errors and copies of everyone else's wildly wrong and obsolete garbage. Lists priced per record. Aha! Adtech is burning down the web and everyone's CPU with all that precise tracking and ML targeting that tells them nothing. Priced per click. How surprising. When Google and Facebook opened some of their profiles to be looked at, maybe 5 years or so back, Google got every major thought about me wrong - my gender, my age, my interests. As it had with most in the office. The whole myth around precision seems no more than a marketing fairy tale to sell ads and justify tracking, very badly. Peak for advertising being useful was very early web with static page ads, simple keyword ads on search, and the odd site sponsorship. Oh, and "customers also bought" on Amazon, that worked well for books and CDs, but doesn't work at all for the 499 other categories they now sell. These days I block everything - JS, uBlock, PiHole. I think there's 10 or 20 sites allowed a little JS, and the odd reluctant exception for bloody hateful reCaptcha. The web's speed is lovely again. Haven't seen a web ad for years - except the odd one or two at work.