Targeted Advertising: Moore's Law Redux?
Moore's Law postulates that the number of transistors in a dense integrated circuit doubles every two years. This principle, first stated by Gordon Moore (then-director of R&D at Fairchild Semiconductor) in 1965, has proven prescient and a good indicator of everything from Internet proliferation to semiconductor supply chains.
Can we interpolate Moore's Law to other digital phenomena? Let's look at targeted advertising, a field well understood by the mid-sixties.
Selling in an analog world
Journalist Vance Packard's 1957 book “The Hidden Persuaders” sold over a million copies and described a dizzying array of persuasion techniques used by American advertisers. “Packard explored advertisers' use of consumer motivational research and other psychological techniques, including depth psychology and subliminal tactics, to manipulate expectations and induce desire for products,” says Wikipedia.
The book describes sophisticated techniques used by ad men to snare consumers' interest and their wallets. For example, car dealerships would promote flashy red convertibles, although those automobiles didn't sell well. The sexy cars would attract male buyers who would then do the sensible thing (under the tutelage of a car salesman) and buy a four-door “family style” sedan — and those sold well. As Packard wrote: “The subconscious is pretty wild and unruly.”
Every WFH team has users who can reroute boxes or set up Pi Holes
Automobile purchasers at that time were a known commodity, so demographics didn't need to be precise or discriminating. Decades later, many advertisers still sought the juicy 18 to 34-year-old sector.
Things have doubled every few years since.
The infinite slice
Let's take a theoretical model using today's precision slicing of datasets to target a specific audience. A CDO wanting more info on his WFH team wants a precise demographic: ages 21-32 with specific language and geographical requirements. How to focus on a group whose parameters match those of your workforce?
A data broker should be able to sell you that package. Full datasets on consumers are valuable. But those datasets must be section-able along precise lines to sell to our CDO.
“Smart” TVs
Analysts of Packard's era never thought of televisions as data gathering tools. But today, the snooping abilities of “smart TVs” are well known.
Earlier this century, Korean manufacturers pioneered techniques to manufacture flat panels at scale. Cheaper displays jackhammered prices into the ground, and a massive flat-screen can be had at white-box shops at near cost.
The real profit is in data collection
How then do manufacturers monetize these units? Simple: the sets themselves are loss leaders. The real profit is in data collection.
My new TV is haunted
Recently, a discussion on the Tech subreddit focused on an article by ZDNet's Senior Technology Editor Jason Perlow titled: “I spent USD3,000 on a Samsung Smart TV — and all I got were ads and unwanted content.” Perlow was not impressed with the Samsung he bought in February 2022.
“You spend this kind of money, cut cords, and pay extra for ad-free services,” wrote Perlow. “But ads are still forced down your throat in the TV UX.”
Readers of the subreddit sympathized. “I connected my TV to the internet, and one day I turned it on, and it made me watch a 10-second ad for orange juice before I could see anything else,” wrote u/DoctorWalrusMD.
Another user with the handle Huge_Nebula_3549 wrote: “We have one Samsung that relentlessly demanded to be connected to a router. It would throw up banners on the screen regardless of the input. So we have a cheap Wi-Fi router that it’s connected to, but the router isn’t connected to the ISP. Just plugged in, in a drawer with its little red lights flashing to infinity.”
More than a few Netizens suggested never connecting the Samsung sets to the Internet at all, using a separate streaming box, or deploying a Pi Hole as a filter. And the inevitable advice to purchase an LG unit instead. But cherry-picking vendors is a shell game.
Another data intrusion, another chaebol
Chris Welch writing for The Verge, titled a 2021 article: “I guess I have to watch ads everywhere on my USD1,500 LG TV now.” “I was updating the streaming apps on my 2020 LG CX OLED TV,” wrote Welch, when “out of nowhere, I saw (and heard) an ad for [vendor redacted] start playing in the lower-left corner. It autoplayed with sound without any action on my part.”
His reaction to this “pretty egregious” intrusion was far from mild. “A random, full-on commercial just popping up in LG’s app store? Is there no escape from this stuff? We’re just going to cram ads into every corner of a TV’s software, huh?”
Data farming
The low price of today's spiffy flat-screens is partially subsidized by the data they harvest. A peek at the Samsung Ads website in the U.S. unveils an advertising mindset that Packard would have understood: “We utilize data technology called ACR, or automatic content recognition, that captures everything happening on the glass and enables consumers a more customized content discovery experience across millions of Samsung Smart TVs.”
Not all consumers are content with that “customized content discovery experience.” And every WFH team contains a few users who know how to reroute boxes or set up Pi Holes.
The lesson of pushback on smart TV ads: users are particular about their in-home experience, and CDOs must remain aware of this when setting parameters for such environments.
Stefan Hammond is a contributing editor to CDOTrends. Best practices, the IoT, payment gateways, robotics, and the ongoing battle against cyberpirates pique his interest. You can reach him at [email protected].
Image credit: iStockphoto/gorodenkoff
