· 5 min read
Paid vs organic when search sends fewer clicks than it used to
If a top ranking now sends 61% fewer clicks, the arithmetic that made organic obviously cheaper than paid has moved. Working it through, with the new numbers.
The paid vs organic argument has always had the same shape: organic traffic is free but slow, paid traffic is instant but rented, and over a long enough horizon the compounding asset beats the metered one. That argument depended on a number that has moved. If a first-place ranking sends materially fewer visitors than it did two years ago, the payback period on every hour of content work stretches, and the comparison deserves redoing rather than repeating.
The number that changed
Reporting through 2026 puts the scale of it beyond argument. Search Engine Land's analysis found search clicks down about 42% from baseline, with a 61% drop in click-through rate on queries where an AI Overview appears, and zero-click behaviour sitting near 69% of all queries.
Those three figures measure different things and it is worth keeping them apart. The 69% is about queries, most of which never sent a click to anybody. The 61% is the one that hits a specific page: it is what happens to your click-through rate when an Overview sits above your result. The 42% is the aggregate effect across a site.
The practical translation: ranking first is still valuable, but "first" now buys roughly the traffic that third or fourth used to. Nothing about the work of ranking got easier.
Redoing the comparison
Take a page that took ten hours to research, write and edit. At a nominal $60/hour that is $600 of cost, whether or not money changed hands.
Under the old numbers, suppose that page ranks and sends 100 visitors a month. Payback against a paid channel at $2 a click is 3 months, and everything after is upside. That is the case for content, and it was a good case.
Now apply the 61% haircut to a query with an Overview on it. The same page sends 39 visitors a month. Payback stretches to about 7.7 months, and it stretches before you account for the risk that the Overview's presence changes again next quarter. The asset still compounds. It just compounds from a lower base, over a horizon long enough that a young company may not be around to collect.
Set that against a bought placement, and the comparison is no longer lopsided. A paid slot has no compounding — when it ends, it ends. What it has is a known denominator: you can see what it delivered, this week, without a forecast. When the organic forecast has just been revised down 61% and may be revised again, the value of knowing goes up sharply relative to the value of compounding.
Where each still wins outright
Organic still wins for queries with commercial intent that Overviews handle badly, for anything where the searcher wants to compare several options rather than take one answer, and for the long tail — specific, low-volume, high-intent queries that no Overview bothers to summarise. It also wins for anything you would have written anyway. The marginal cost of publishing what you already know is near zero, and near-zero cost survives any traffic haircut.
Paid wins when you need an answer this week rather than this quarter, when you are testing which message deserves the ten hours of writing, and when the channel publishes what it delivered. That last one is not a general property of paid media — most channels report to the buyer, privately, after the spend, which is a different thing from publishing.
Neither wins if you cannot measure the outcome, and that failure mode is now more common than it was, because the attribution layer got worse at the same time the click volume got worse.
What to do this quarter
Three moves, in order.
Check which of your pages actually face an Overview. The 61% figure applies to those queries only. Some of your pages are unaffected and are still doing their old job; treating the whole site as if it took the haircut will lead you to abandon work that is fine.
Test messages before you write. If you are about to spend ten hours on a page, spend an hour first putting the headline in front of cold traffic and see whether anyone clicks it. A message that cannot earn a click on a paid placement will not earn one from a search result either — and that test now costs less than a tenth of what the page costs. This is the argument for the one day ad campaign as a screening step rather than a growth channel.
Stop treating the two as a budget split. The useful framing is not "what percentage goes to paid". It is: paid buys you information quickly, organic buys you an asset slowly, and information should come first because it tells you which asset to build.
Our own board sits in the "paid, and it publishes what it delivered" corner — two twelve-hour slots a day, the first brand into an empty one pays nothing, and the opens and click-throughs each slot produced are public on the front page and on each brand's own page. That is ours, so weigh it accordingly; the honest limitation is scale, and it is a young board that says so with numbers rather than adjectives.
The uncomfortable conclusion
The strongest version of the case for organic was never "it is free". It was "it is free and predictable enough to plan around". The second half of that has weakened, and the first half was always an accounting fiction that ignored the writing.
That does not make content a bad investment. It makes it a slower, riskier investment than it was two years ago — the same shift that makes a twelve-hour ad slot's cost per visit worth computing honestly, which means the bar for starting one should be higher and the case for testing the idea cheaply first is stronger than it has ever been.