The default is the lever

When Google announced an ad blocker inside Chrome, the obvious reading was that it would quietly spare Google’s own ads.

It’s a good guess. I think it’s wrong. Google will block their own ads and the reason has nothing to do with restraint. They don’t need the exception because they have something that works better and costs them nothing to use.

Start with what the exception would cost. A browser holding two thirds of the market, shipping a blocker that mysteriously classifies its owner’s inventory as acceptable, is a story that writes itself and then runs for years afterwards every time anybody writes about online advertising. The amount of trust lost would be immense and the gain is a slice of inventory they already dominate and nobody at that company is stupid enough to take that trade. It isn’t a moral position. It’s arithmetic. The arithmetic is not close.

Then look at what you have instead if you are the browser vendor. You control what the setting does when nobody touches it, which turns out to be almost everybody who will ever use the thing you shipped. That’s the whole game. It’s worth more than any blocklist could ever be, because the overwhelming majority of people will never open a settings panel in their entire life and whatever the browser does without being asked is therefore what the web does. Whatever the browser does out of the box is what the web does.

So the interesting question was never which ads get blocked. It’s what happens to the ones that don’t and who gets to decide which category any given ad falls into in the first place, because that decision is made once and then applies to everything afterwards without further argument. A blocker that ships switched off by default doesn’t need to protect anyone’s inventory. A blocker that ships switched on, with a definition of “acceptable” written by the largest seller of acceptable ads, doesn’t need to either. In both cases the outcome is set before any individual ad is judged, by a decision most users will never know was made.

This is the thing that’s easy to miss when you’re watching for the obvious corruption. You look for the thumb on the scale and you find the scale is fine because the thumb is on the question of who gets weighed. Blocklists are visible, arguable and auditable, which is exactly why they attract every ounce of the scrutiny that gets spent on this. Defaults are none of those things and they quietly move more traffic than any blocklist ever written.

Once you’re looking for it you can see the same shape more or less everywhere, in places that have nothing to do with advertising at all. The cookie banner that defaults to accept, the search engine that ships pre-selected, the analytics that’s on unless you turn it off, the sharing setting that starts at public. In every case there’s a visible mechanism that gets all the argument and a default that quietly does the actual work. The argument about the mechanism is genuine and it’s also a bit of a distraction from the part that decides the outcome.

Which is why “will they exempt themselves” was the wrong thing to be worried about, and it’s a question about the mechanism and the mechanism was always going to be clean. If you want to know what a browser vendor is going to do to the web, don’t read the blocklist. Find out what the checkbox says before anybody touches it and how hard it is to find.

None of that requires anyone to behave badly and that is what makes it durable. A blocklist has to be defended in public by somebody who can be asked why a particular ad is on it. A default gets set once, by a product decision nobody outside the company will ever read, and then it holds until someone notices it and cares enough to go and change it, which for the overwhelming majority of people is never. The mechanism gets all of the scrutiny for the simple reason that the mechanism is the part you can actually see, and you will spend your attention on the visible thing every time unless you make a deliberate effort not to.

So when a browser vendor tells you what its blocker will and will not stop, that is genuinely worth reading and it is not the important document. The important one is the spec for the first run experience and nobody publishes that.