Empirical analysis of behavioral advertising finds that surveillance makes ads only 4% more profitable for media companies

Originally published at: https://boingboing.net/2019/06/04/all-that-for-4-pct.html


The executives and shareholders aren’t going to leave that extra 4% on the table. It’s not like the externalities of the business model that delivers it will ever affect them, amirite?


Companies have removed employee benefits, like removing free coffee, with monetary results that overall were statistically insignificant (much less than .01%); so, 4% sounds like a gold mine of wealth to be exploited!


By now I could be easily convinced that targeted advertising has just been a pretense for this all along. what a spectacle.


I notice that FireFox isn’t on the list.
Probably because all of us FF users install ad-blockers and the like.


Three observations.

  1. I’m surprised it’s that high.

  2. The sentence “surveillance makes ads only 4% more profitable for media companies” really means “surveillance is incredibly effective from the perspective of media companies.”

In the meantime, surveillance imposes enormous external costs on society

  1. There is no line on business budgets for external costs.

Harry Shearer had an interview with the author of Surveillance Capitalism, Shoshana Zuboff. She goes into some detail about how gathering this data is an end unto itself, whether it’s useful to advertisers or not.


Meanwhile outfits like DuckDuckGo that optimize by not tracking and cutting costs (they’re based in Philly, not the Bay, and have a remote workforce) stand a serious chance of eating Google’s lunch in the same way Vanguard has cornered the investment market by offering low cost index funds.

1 Like

Nice headline, ya godless commie! It should read:

Empirical analysis of behavioral advertising finds that surveillance actually makes ads 4% more profitable for media companies

Hell, if eating babies made ads 2% more profitable, three guesses as to what the new Madison Avenue dining trend would be.


This is no surprise, as often the ads i do see are late, so the other month i purchased a new cordless drill, i spent about 20 mins googling it and visiting other web sites, to see if the deal i had was the cheapest deal which it was so, i went and purchased it, but on line advertisers dont know this, and 4 weeks later i am still seeing adverts for a drill i purchased and am not really going to be a second drill…

Is there any info on what DuckDuckGo’s business model actually is?

(A lazy probe around their site left me… still uncertain. They might be supported by donations?)

I totally buy it.

Humans are not so easy to push into behaviours as marketers like to pretend. Convincing someone to spend in-game coin is not the same as having a single mother part with hard-found cash that needs to feed her kids.

Techery can be a bit spoiled and full of bullshit. Development teams isolated in a psychological island paradise of wealthy software buggers will not understand the majority of people.

So 4% - yeah, sounds right.

How to increase it? Oh, addiction, bullying, gouging.

I think that the 4% increase could be a statstical error. I normally take the paper advertising of supermarkets because I am interested on special offers on groceries, and if there’s a special offer on groceries I need or I normally use, I’ll go in a specific supermarket. If I am interested on buying CD I’ll normally search about hese itemo on some sites. The targetting will work the same if it will be site-based. In television if there’s a bike race, and advert on a bicycle or a a saddle will work more than one on beer. I remember in the '90 buying computer magazine to read about computers, and these magazines were fille with computer-related ads.

It’s more useful to find the correct dempgraphics.

1 Like
1 Like

So let’s ramp up surveillance by 200% and we can get another 8% out of it! Woohoo! Hold my coffee, I am out to buy some stock…

1 Like

Let me just re-write that headline from the media company perspective for you:

Empirical analysis of behavioral advertising finds that surveillance makes ads 4% more profitable for media companies !!!

I have full confidence that they’d steal our organs in our sleep if the ROI was decent enough.


Considering that the entire purpose of marketing and advertising is to manipulate people, the fact that they haven’t literally taken over the world is proof that it’s probably not as effective as they claim.

1 Like

I don’t know why Firefox isn’t in Table 1, they certainly have some data about firefox usage:

The probability of a cookie being associated to the user seems to significantly relate to technical aspects having to do with the type and model of the device employed by the user (Table 5). For example, users that use Chrome, Explorer, and Firefox are significantly more likely to have a cookie associated, compared to those who use Safari. Users who are browsing using a mobile device (either mobile phone or tablet) are less likely to have a cookie associated than users browsing using a desktop or laptop. This may reflect the fact that the cookie tracking abilities are limited in mobile environments, specifically within apps (IAB, 2013). The geographical location matters as well, as it can capture differences in laws and regulations; for example, European IP addresses are less likely to have a cookie associated, relative to US IP addresses. Similarly, Canadian, Asian, and Australian IP addresses are less likely to have a cookie associated, relative to US IP addresses. Instead, features more specific to the users, such as the topic browsed (sports, motors, fashion, etc.), do not seem to be associated with the likelihood of having a cookie or not. The treatment model also includes date and time fixed effects, website fixed effects and other features (such as browser version, operating system version, mobile phone type, mobile phone carrier, and topics browsed by the user).

They speculate that there are fewer cookies on Safari because Safari blocks third party cookies by default.

About 73% of the ads shown on a Safari browser do not have a cookie associated, whereas on Chrome this is the case about 17% of the time. The difference is probably due to different default tracking settings across the two browsers, with Safari impeding, by default, third-party tracking cookies being set on the user’s machine (the user has to explicitly allow the usage of third-party cookies) (TheVerge, 2013).

The authors sort of deal with those points:

We hit peak innovation in most industries a while back. Now the only innovations left are marginal at best, and come at enormous costs to public institutions and the public trust at large. This is the part of the story where market capitalism resorts to attempting to squeeze blood from a stone.