The guys from the first round of online advertising agencies that went bust around 2000-1 could probably confirm or dispute that for us.
I more-or-less agree with your main point (which is to say, analytics are a desperate, losing attempt to make online advertising work better, because it doesn't work very well) but there are many other things in the post that I think are confusing (I don't think there are spreadsheets; I don't think the information's being resold, exactly, so much as used to sell other things (ad space, products for collecting similar such information). Anyhow I think what you're really predicting is imminent collapse for google, which, could be, could be.
It's probably an SQL-able dataset, but it could be put into a whole bunch of spreadsheets Mr. Pedantic Programer.
imminent collapse for google, which, could be, could be.
Sometimes it freaks me out that my cell phone operating system was designed by the same company I use to search for stupid things on the internet. But, really, my cows.
I realize that you're using Etsy merely as an example, but they do use a simplified form of personalized data in conjunction with Facebook. If you accept the Etsy app for Facebook, Etsy will pull all the interests that your friends specified in their profiles, match the interests to products in their store and provide you with the suggested products. I have no idea if there is any money being exchanged between Etsy and Facebook, but it's a clever idea that should result in increased sales for Etsy.
I also wonder if the system that paid for all of our television for the last half century made economic sense.
I have no idea if there is any money being exchanged between Etsy and Facebook, but it's a clever idea that should result in increased sales for Etsy. swell example of the creepy invasion-of-privacy vibe that these targeted recommendation engines always give off.
On the other hand, I'm now tempted to list "plush vagina wall hangings" among my facebook interests just in case anybody installed that app.
Except isn't Etsy all people reselling cheap factory-manufactured knicknacks for a 1200% markup, now?
No system that gave people money for making Mr. Belvedere can be morally justified not matter how well it worked economically.
7: You can get a knitted, crotchless body suit. Where else can you get that?
6: well, that's the thing with adverstising. It certainly does seem to work on some level, but it's never been something you could meaningfully quantify. Because internet advertising necessarily has less reach and impact than TV advertising (or promotional stunts or sponsoring research labs or whatever Nike's up to now), the added value that they lit upon was analytics (which is to say quantifiability). Excep, since the effect of advertising isn't really meaningfully quantifiable, they've boxed themselves into a corner where they have to keep trying to invent metrics that will believably show that advertising is working, when that's basically an impossible problem.
I've seen an hypothesis floated that an alternative way of seeing the relative failure of internet advertising is that it's actually no less successful than other forms of advertising; it's just that we can quantify that it fails in a much more precise way than we can with traditional media ads. Not sure how you would test this.
I agree that there's a lot of swirling bullshit. But facebook definitely knows which reality TV celebrity is most popular this week among wealthy 40-year old single women, and which among 16-year old boys. That information has some value to publishers.
The target audience for advertising is not SWPL cheapskates, vendors of goods are interested in pushing their ads for jewelry or "medicine" at the wealthiest customers, so zipcode-level ad control is extremely valuable. For seervices, it is not clear there is a bubble. The most expensive search terms on google look pretty productive to me. The word mesothelioma for example is very expensive, I think wrapped up into category 4.
Lastly, etsy hasn't worked all that well for me. The people I know who make money from their art have either closed their etsy or use it as a storefront where to direct interest generated by other means. I know this was just an example, chosen as an instance of generic small business, but etsy itself as far as I can tell is not generating sales. Probably part of the problem is that handblown glass takes expensive natural gas and so is expensive compared to machine-made or minimally worked glass, let alone other decorative materials.
6: Exactly what I have wondered, starting when I was about 6 or so--it just did not seem possible that ads could actually pay for TV shows. Part of the answer is that attention really is the scarce resource in the commercial world.
Probably part of the problem is that handblown glass takes expensive natural gas...
People keep fracking around here, so it should get cheaper.
Innovative British startup is snooping on you at the mall! I wonder how many of their employees are ex-14(EW) Signals Regiment...
No, etsy has at the very least many people who do nice watercolors.
http://www.etsy.com/listing/67878041/new-england-winter-scape-no53-original
http://www.etsy.com/listing/77704038/barn-owl-10x8-inch-original-painting
OK, or glass
http://www.etsy.com/listing/68137734/vases-with-a-pinch-and-a-twist
Etsy will pull all the interests that your friends specified in their profiles, match the interests to products in their store and provide you with the suggested products
Well, if there were Heiner Müller items for sale on etsy, I would definitely want to know about them.
Brecht does yield a few hits, including this homemade Threepenny Opera painting. What's more Threepenny Opera than skulls, hearts, and daggers?
This topic -- the amount of value added created by targeted advertising, came up in conversation at Thanksgiving, though none of us had an answer.
The more surprising part of that conversation was being reminded of how non-bloggy most people are. I have two cousins who work in computer related fields (one a programmer, one a consultant) and I asked them a bit about their feelings about privacy and the efforts by companies like google and facebook to create a single profile of people's online identity (at least internally, for their own tracking purposes). It quickly became clear that neither of them had any intuitive sense of why people would want to have an (one or more) online identities which was distinct from their real-life identity. That just sounded a little non-sensical (or, at least, fringe) to both of them.
It just reminded me much unfogged has influenced my sense of the world (both on and offline) since I take for granted the idea that independent online identies are valuable.
The quantifiability of advertising is killing Buck -- since the late eighties, he's been writing a series of advertising-driven computer newsletters. Advertising money used to flow freely, when no one could measure what the effect of it was. Now that people can count click-throughs, they're very reluctant to spend.
Now that people can count click-throughs
As poor a measure of ad influence as has ever been designed. Most of what I talked in 10 is a desperate effort to recover from the horribly foolish decision to price ads based on click-throughs.
9: I was informed recently that the friend who had me knit him a loincloth over this NWSFish pattern had it thrown away by his new wife, which I suppose I can understand. I refused to do any fittings or whatnot, so it may not have been the right size anyhow.
18: ...neither of them had any intuitive sense of why people would want to have an (one or more) online identities which was distinct from their real-life identity.
I think this is very common. It bothers me because it's driving discussions of online privacy in the larger world outside blogdom. People who've seen online discussions turn into real-world harassment have a very different take on it, but we're a pretty small minority. Also a lot of people simply don't have controversial opinions or even any clear opinion at all on controversial subjects.
The thing about targeted advertising is that its basically cannibalizing non-targeted advertising. Its not like the amount of money I spend purchasing consumer crap is being increased. It might be shifted around due to targeting, but its not actually making a difference in the total amount of money that marketers are able to squeeze out of me via marketing. So, if more of those dollars are going to Google, I think that just means less are going to TV advertising and newspaper display ads.
Which means I don't think its a bubble. Or not much of one. Its just a redistribution of existing spending patterns.
And, you know, some of the money is being spent on rather good search algorithms and big data centres. Rather than just coke'n'hookers like in traditional advertising.
24: So I guess in 40 years we won't have any good TV shows about the crazy, rollicking lives and careers of the people who worked in internet advertising?
Pretty sure internet advertising people still spend plenty of money on coke, at least.
I love Etsy for handmade goods. I'm terrible at knitting, so I've bought gloves in the shape of owls and crocheted cockatiels from them. The cockatiels are particularly awesome because I sent the seller pictures of my birds and she knitted the cockatiels to have the same patterns. I also bought my husband a tiny T-Rex in a bottle, because it was so unique.
I've seen an hypothesis floated that an alternative way of seeing the relative failure of internet advertising is that it's actually no less successful than other forms of advertising; it's just that we can quantify that it fails in a much more precise way than we can with traditional media ads. Not sure how you would test this.
This is basically my take, namely that the long term problem (from an online content provider's point of view anyway) isn't that people don't pay enough for online ads, it's that they pay too much for traditional ads. At least for certain types of advertisers/advertisements. Eventually the two rates will converge, but more from traditional ad rates coming down than internet ad rates going up.
But that's entirely speculation from a journalist's perspective.
11/23/28 is my take as well - money moving out of forms of advertising with next-to-no metrics into new forms with more metrics (as mentioned, metrics of dubious value). As someone who benefits from this trend and gets to do cool things as a result, I hope it continues.
November 23, 2028 is a long ways off.
You'll find papers from Google and Yahoo evaluating online ad effectiveness. No prizes for guessing the results. They have stats and stuff to back up their claims.
I looked for an example but couldn't find one, though I have read such papers in the past.
November 23, 2028 is a long ways off.
Thank god, seeing as my eldest will turn a horrendously old 32 that day.
(JRoth will be pretty old too!)
If JRoth were to travel at near the speed of light until then, he won't be as old as the rest of us.
I'm sure Knecht or someone more expert can comment more intelligently on this, but I thought that the value of Google and Facebook wasn't just based on assumptions about the power of targeted advertising, but on the ability to collect data that will inform better marketing, which is much more valuable; i.e. not just selling a product, but knowing which products to make in the first place and how to distribute them. And a sense that eventually Facebook and Google will themselves become distribution channels.
That still smells like a bubble to me, especially since the two companies themselves don't seem great at anticipating markets, but maybe it's a slightly different kind of bubble.
I think you're right that the stock proces of google and facebook have a lot of vague expectations about the value of data aggregation built into them, but aren't they partly (laregly?) based on good old fashioned things like the fact that both companies (much moreso google) are currently legitimate cash cows and also the fact that both companies (much moreso facebook) are becoming so imbedded in people's lives that they're developing monopoly-like market power?
Yeah, by "become distribution channels" I basically meant "become so entwined in people's lives that they can exercise monopoly control and sell us whatever they want." Of course this involves changing the business plan from "don't be evil" (the information gathering stage) to "be extraordinarily evil."
31: Here's one paper: http://mfi.uchicago.edu/events/20111028_experiments/papers/lewis_rao_hardability.pdf
When I ask folks at Google and Yahoo about this, I get a few answers. Besides pointers to research like that, they'll say that online advertising is much more cost effective than traditional advertising, so, regardless of its effectiveness, it's a better deal.
Also, they're moving into more traditional ads in various ways (selling display ads in online venues, using the technology they use to sell ads for online venues to auction ads for TV, etc.).
I don't know how it works at FB, but at Google/yahoo, I'm told the data isn't re-sold; it's used so that people who buy ads get their ads displayed to people who want to see them. Etsy shops can afford it because all they do is buy some relevant keyword, and then pay a few cents per click. Some keywords (e.g., remortgage or vioxx class action) will sell for tens of dollars per click. It's hard to believe any of those aren't working, because it's trivial to tell if the ad campaign is paying off, as long as you're not a huge corporation that can expect that people might have heard of your product anyway.
Pretty sure internet advertising people still spend plenty of money on coke, at least.
I thought the traditional choice of internet people was Mountain Dew.
Cocaine and Mountain Dew go together just fine.
And by "Mountain Dew" I meant ScotchGard.
I got a moderately puzzling targeted ad on FB the other day for "The first female-fronted punk rock band from Greece making its US debut."
The ad was fairly well targeted. I love punk and things that are female-fronted. And between philosophy and debt crises I type the word "Greek" on facebook a lot.
But how many of these descriptors have to apply to the band before the claim is true. Surely they cannot be the first Greek punk band, or even the first female-fronted Greek punk band. Perhaps they are just the first female-fronted Greek punk band to play the US?
Also, if anyone has any recommendations for female fronted Greek punk bands, I'd love to hear them.
One of my FB friends is a promoter who I'd swear brought over a Greek lady punk rock band last summer. I'll have to check my email. (The guy is a Johnnie but I don't think you know him.)
The younger daughter in the movie "Dogtooth" is the female frontress of a female-fronted Greek punk band that also contains the movie's director.
Meanwhile, a rare instance of extremely good news.
Rembetika has a protopunk sensibility. Σωτηρία Μπέλλου is pretty good.
45: assuming it doesn't lead to the SEC just quietly dropping the case.
I used to be into female-fronted Greek punk rock bands, but then they got all trendy when they started coming to the US. I've since moved on to androgynous Macedonian ska.
48.last: oh yeah, Berlin Alexanderskatz. Love them.
I thought the traditional choice of internet people was Mountain Dew.
Please. It's Jolt.
49 is so terrible that I suspect collusion. Sifu is a ringer!
It's probably an SQL-able dataset, but it could be put into a whole bunch of spreadsheets Mr. Pedantic Programer.
My understanding is that the big sites with lots of data are running some kind of NoSQL - traditional relational databases can't scale like that.
The guy is a Johnnie
Is this internet privacy shorthand or do you really call yourselves that? If the latter, I sort of want to judge it harshly, but the nickname used at my own DFH alma mater is worse.
They call themselves that. It doesn't come off unbearable in person, I don't think.
Does anyone at Yale call her/himself an Eli? Or is that strictly limited to the generation that always wore hats and the NYT crossword puzzle?
Yep, LB is right. We really say it -- and not even in a mannered, goofy way. Totally matter-of-fact.
the generation that always wore hats and the NYT crossword puzzle?
A striking costume, if draughty.
Graduates of my first-attended institution of higher learning seldom unironically call themselves Massholes.
58: English: one comic misadventure after another.