There is, supposedly, a deal. You get the article, the recipe, the league table; in exchange, the page gets a few seconds of your attention, which it sells to whoever bids the most. Content for eyeballs — the founding bargain of the free web. It has one structural flaw, and the flaw is load-bearing: nobody ever asked you.
No terms were presented and no box was ticked. The deal is a custom that hardened into an entitlement, and while you were reading it quietly amended itself. Attention alone stopped being the price years ago; the going rate became attention plus a dossier — identifiers, trackers, and an auction for your eyeballs that clears in about a hundred milliseconds, before the headline has finished rendering. An ad blocker is what it looks like when the other party to a contract they never signed finally reads it.
What a blocked ad actually costs
The mechanics are worth being precise about, because they are brutal. With a blocker running, the ad slot simply never loads. No bid request goes out, no auction fires, no impression is recorded, no money moves. A blocked impression is not a discounted impression; it is zero. There is no partial credit, and for most publishers no fallback — the hosting bill, the CDN and the writer's invoice all arrive regardless. The reader consumed the product and the till stayed shut.
3 in 10
This is no longer marginal behaviour. Roughly three in ten internet users tell researchers they block ads at least some of the time,1 and the industry's own census — kept, as it happens, by an ad-filtering company — counts more than 900 million active ad-blocking users worldwide.2 For a mid-sized publisher that is not leakage. That is three readers in ten walking straight past the till.
Here is the part that should bother you more than it bothers the industry: the pain lands on the wrong desks. Blocking does not discriminate by behaviour; it discriminates by audience. Technically fluent readers — the ones a careful publisher most wants — block at the highest rates, while the made-for-advertising mills, built to farm whoever arrives unarmoured, barely notice. Punish ads as a category and you starve the restrained sites first, because the shameless ones were never relying on your goodwill.
A blocked impression is not a discounted impression. It is zero — and the writer's invoice arrives regardless.
The toll-booth in the middle
Now for the blockers' own hands, which are not spotless. The best-known filtering lineage — Adblock Plus and the Acceptable Ads programme run by its owner, eyeo — does not in fact block everything. Ads that meet the programme's standards for size, placement and labelling pass through by default. Being whitelisted is free for most participating sites, but large entities — those gaining more than ten million extra impressions a month from being waved through — pay eyeo a licensing fee of roughly thirty per cent of the additional revenue the whitelisting creates.3
Whatever that arrangement is, it is not neutral self-defence. A blocker that charges the advertising industry for safe passage has become a market-maker in the thing it blocks, with its users' attention as the inventory. You installed a shield; it turns out to moonlight as a toll-booth. None of this makes blocking wrong — but it should retire the fantasy that the tools are pure and only the publishers are compromised.
The case for the shield
Because the case for blocking is real, and it is not about being cheap. In March 2016, visitors to the New York Times, the BBC, MSN, AOL and Newsweek were served ads rigged to install ransomware, through the sites' own ad slots. The publications were not hacked; their ad networks were tricked into delivering the payload for them.4 When a page might quietly try to encrypt your hard drive, refusing to load its ads is not free-riding. It is the instinct that makes you lock a door.
The cost argument is nearly as strong. When the New York Times measured the mobile pages of fifty news sites in 2015, advertising and its scripts accounted for more than half the data on many of them; at the extreme, one site's ads weighed almost four times its journalism.5 On a metered connection, the reader is paying — in the plainest, per-megabyte sense — to be advertised at.
And beneath both sits the argument the industry least wants to hear: readers did not break the deal first. Content-for-attention became content-for-attention-plus-surveillance without anyone renegotiating; the tracking was an amendment nobody initialled. When one side rewrites a contract unilaterally, the other side's exit is not betrayal. It is the only clause they were ever given.
Blocking like you mean it
So two things are true at once, and I would rather hold both than blend them into mush. Blocking is rational self-defence against a supply chain that has delivered ransomware and eaten data plans. And blocking as an unthinking default collectively starves exactly the publishers behaving best, while the worst barely feel it. The coherent position is not block everything, and it is not block nothing. It is block by default, and whitelist like you mean it: per site, deliberately, for publishers who run restrained ads, skip the twelve-tracker payload and label their sponsorships. Whitelisting is the reader finally at the negotiating table — the deal that was never offered becomes one you extend, site by site, to whoever earns it.
The deal that was never offered becomes one you extend, site by site, to whoever earns it.
You can run the experiment here. This blog is ad-funded and says so — the machine pays for the writing about the machine. Visit with a blocker on and the Exhibit on the landing page degrades to four labelled empty frames: the auction never fires, the frames stay politely blank, and the essays load anyway. I am not going to ask you to whitelist this site; the empty frames make the argument more elegantly than the ads do. But decide, rather than default. The deal was never negotiated. Now, one site at a time, it can be.