There are two kinds of advertising

  1. Contextual
  2. Targeted

Not to bury my point: contextual is the normal, good, fair, effective type of advertising. Targeted is the creepy, resource intensive, privacy invading, and, (?!?!?!!) not particularly effective type of advertising. We’ll get there.

A billboard is contextual advertising. You’re driving on the highway and are told there is a McDonalds in 7 miles. The human condition + geography are the context.

An ad for web hosting on a blog about web development is contextual advertising. I lived this for well over a decade. Web developers are the customer for web hosting so reaching them on a blog about what they do is perfectly contextual. I sold ads to companies that wanted to reach web designers and developers, and I still do on CodePen.

Here’s more contextual advertising. Lifeline (“I’ve fallen and I can’t get up!”) runs and ad on Wheel of Fortune (old people watch The Wheel and need Lifeline). Purina sponsors the local pet parade (there are lots of dog owners there who make a choice about what dog food to buy). An affordable e-bike advertises on busses (people there might be into a commuting alternative). It tends to work.

Targeted advertising, the other kind, is an advent of the digital age. Its the kind that says I only want to show this ad to woman aged 36-44 in Baltimore using an iPhone who have previously purchased shoes online, have stopped to look at at least 3 shoe advertisements in the last week.

How do they know that stuff? I mean — if it’s Facebook you’re just telling them lol. But generally: it’s tracking tech.

Go to The Verge (just to poke at a site I generally like) without an ad blocker, open up the Network panel in DevTools and just let ‘er rip. I’m seeing 400+ requests. That’s tracking at work. You can even just sit there and watch it continue to make requests over time, even while you’re doing nothing. JavaScript is whirring, soaking up whatever data it can, setting cookies, and blasting data along with your precious IP address to god-knows-where. All those requests are slowing down the site, costing you bandwidth, laughing at your privacy, and causing legislation that at least you have to click a giant content-blocking banner with a “yes, this is fine.” button.

But at least it makes advertising better, right? At least it ensures that advertising is super relevant to you, much better than contextual, right?! At least companies are getting a ton of bang for the buck, right?!?!

No:

… three deep, recent studies that show: broad reach beats targeting for incremental growth; that the cost of targeting outweighs the return; and that second and third party data does not outperform a random sample. First party data does beat the random sample – but contextual ads massively outperform even first party data. And they are much, much cheaper.

Those three studies:

  1. Overwhelming targeting options”, 2023, Ahmadi, Nabout, Skiera, Maleki and Fladenhofer 
  2. “Is it wise to prioritise #Reach over #Conversions?” from Rikard Wiberg at PACE
  3. Is first- or third-party audience data more effective for reaching the ‘right’ customers?, Nico Neumann, Catherine E. Tucker, Kumar Subramanyam, and John Marshall

Effectiveness testing data is evidence that contextual ads, sorry, the normal, good, fair type cost less and perform better. And as web developers, we know they are better for performance. And as humans, we know they are more ethical.

So what in the smurfing smurf is going on here?

It’s a follow-the-money situation. It’s likely as simple as selling snake oil to cure anything that ails ya. It feels good to buy advertising that is directly (and only) toward people you deem the most likely to become a customer and buyers are willing to pay for that, even if it’s about as effective as actual snake oil.

Hat tip to Jeremy who blogged about this and pointed at the article covering the studies, which includes an actual challenge:

If targeted advertising is going to get preferential treatment from browser makers, I too would like to see some evidence that it actually works.

Let’s see all that proof that 400+ requests for thirsty ass always-running JavaScript is just what we have to do to make advertising good.

This is all in the wake of Chrome saying uhhhhh ya know how we were going to deprecate third-party cookies, a staple of tracking technology? Just kidding. Not only are we not going to do that, we’re going to create new technology on top of it to help tracking more.

Google is uniquely suited to shut all this down and help with contextual advertising what with, ya know, a powerful-if-monopolistic search engine that knows an awful lot about the context of essentially all websites. But instead, the muscle and influence gets behind tracking and targeting.

Thoughts? Email me or comment below. Also CodePen PRO is quite a deal. 🙏

One response to “There are two kinds of advertising”

  1. Nicolas says:

    As an aside: in the physical word, the most common type of advertising is still “Aimless advertising”, i.e. giant prints of foldable telephones on the scaffolding at the main square, metro stations walls entirely covered in pictures of oat milk cartons, big screens blinking about plastic surgery on my road to work, latest new bestselling book on the flanks on half of the busses. Very little of physical advertising is actually contextual, and most people seeing it will have no interest in the product being sold.

    I do agree about contextual on the web being reasonable, while targeted advertising is an aberration.

    I’m not sure it’s that much about advertisers feel good about buying it though: I believe it’s mostly because a couple of giant businesses made their entire model around selling targeted advertising, and built entire suites of free products made only to collect data from their users. If the truth around targeted advertising was to be common knowledge, these two or three companies will have to pop out of existence, and they won’t let that happen.

Leave a Reply to Nicolas Cancel reply

Your email address will not be published. Required fields are marked *