Nobody blocks advertising

Everyone in advertising has a theory about why people block ads, and every theory is flattering to whoever holds it. Publishers think readers are freeloading. Ad networks think a few bad actors ruined it. Privacy people think it is all about tracking. Each of those is a partial truth that happens to exonerate the person saying it.

There is a way to find out that does not involve asking anyone, and I would rather use it, because people are unreliable narrators of their own irritation. Nobody remembers the specific moment they installed uBlock Origin. They remember being fed up.

So read the filter lists instead.

EasyList and its relatives are public. They are a written record, maintained for over fifteen years by volunteers, of every specific thing enough people objected to that somebody sat down and wrote a rule about it. It is the closest thing the web has to a transcript of the grievance. And when you read it, four categories come out, in roughly this order of how much they drove people to install anything at all.

1. Interruption

The largest body of rules by far is not about advertising as such. It is about things that got in the way. Interstitials you have to dismiss before reading. Countdown timers. Video that starts playing with sound while you are looking at a different tab. Sticky bars that eat a third of a phone screen. Content that reflows twice while you are mid-sentence because a slot loaded late.

This is the one that actually made blockers mainstream, and it has almost nothing to do with commerce. A newspaper full of quarter-page ads has never generated a movement. An article that covers itself with a modal before you reach the second paragraph generates one immediately. The complaint is not "you are selling me something." It is "I came here to read and you stopped me."

2. Weight

The second category is cost, and it is a cost the reader pays. A page with a dozen ad slots is not one extra request. It is an auction: header bidding scripts, several exchanges, identity providers, measurement vendors, each one pulling in the next. It is entirely ordinary for the advertising on a page to outweigh the writing by an order of magnitude.

On a laptop on good wifi you experience that as a slightly janky page. On a mid-range phone on a bad connection you experience it as a site that does not work, and as a data allowance that runs out before the end of the month. A lot of people did not install a blocker for ideological reasons. They installed it because their phone got usable again.

3. Surveillance

Here is the one that everybody in my industry wants to talk about, and it belongs third, not first.

Third is not "unimportant". Third is where it actually sits in why people first installed something -- and it is the category growing fastest, the one that skews hardest towards people who understand what is happening, and the only one of the four that cannot be fixed by being tasteful.

That last part is the whole point, so let me be plain about it. You can fix interruption by not interrupting. You can fix weight by shipping less. Both are choices about restraint, and any publisher can make them tomorrow afternoon.

You cannot fix tracking by being restrained, because tracking is not a property of how the ad behaves. It is a property of where the request goes. A single quiet, tasteful, respectful banner served from a third-party ad host still tells that host which article this person is reading, at what time, from which address, with which browser fingerprint, alongside everything else it has ever seen them read anywhere. The banner can be the most beautifully behaved object on the page and the disclosure is identical.

That is why privacy is the reason that persists after everything else is cleaned up. A publisher who tidies their layout has genuinely fixed reasons one and two. They have not touched reason three, because reason three was never about the advertisement. It was about the introduction.

4. Danger

The smallest category and the hardest to argue with. Programmatic advertising means a publisher does not know, at the moment a page loads, what will appear on it. That is not negligence, it is how the auction works. It also means malicious creatives get served by reputable sites periodically, and have for years.

Once someone has had a fake virus warning take over their browser on a site they trusted, they install a blocker and they never uninstall it. You cannot win that person back with better ad design. They are not making an aesthetic judgment.

What is not on the list

Now the part that publishers, including me, would rather skip.

Read the whole of EasyList and you will not find a rule whose purpose is "this website is trying to make money." There is no filter for commerce. There is no rule against a sponsor being named, against a product being recommended, against a page being paid for. Newsletters carry ads to enormous audiences of people who block everything and nobody minds. Podcasts read sponsor copy out loud to listeners who run Pi-holes.

An ad blocker is not an anti-advertising tool. It is a reaction to a specific bundle of behaviours: interrupt, weigh down, follow, occasionally attack. The industry chose every one of those, deliberately, over about fifteen years, because each one individually made a number go up. The blockers are the bill.

I find this genuinely clarifying rather than depressing, because it means the objection is to conduct, and conduct is changeable. It also means the sympathetic version of the publisher's complaint -- "people block ads, so journalism cannot be funded" -- has the causation backwards. People are not blocking your funding. They are blocking four things you did to them, and you can stop doing all four.

The uncomfortable corollary

If the objection is to conduct rather than to advertising, then anyone who fixes the conduct has a real argument to make. I have a stake here, so read the next paragraph with that in mind.

You do not get to remove the behaviours and then also claim the moral high ground for getting past the tool that existed because of them. Those are separate acts. If your ads make no third-party request, carry no script, follow nobody and interrupt nothing, then the honest description is that you removed the reasons -- not that you defeated the defence. A reader who still wants nothing to do with you must still be able to say no, easily, and the ads must still be labelled as ads. Otherwise you are just the next thing the list gets updated for, and you will deserve it.

There is a simple test I keep coming back to, and it cuts against my own interests often enough that I trust it.

Would this advertisement survive being shown to somebody who has told you, by installing a blocker, that they do not want advertising?

If it would -- if it is a static image, in a fixed space, that loads instantly, phones nobody, and sits there being a picture of something for sale -- then that reader's objection was never really to you. You were collateral damage in a war they were fighting with somebody louder.

If it would not, then the blocking is not censorship and it is not theft. It is feedback. It is the most direct feedback the web has ever produced, delivered by millions of people, at the only moment they were actually offered a choice.