When the cost of the next page exceeds what it returns. Organic search has no cost per click, but it has a real cost per page, and that cost rises as you work through the terms you can win.
Why is organic described as free when it is not?
Because the invoice arrives in a different form. Nobody charges you per visit, so the spending shows up as salaries, writer fees, developer hours and tool subscriptions, and those sit in a budget line that is never divided by the clicks it produced.
Run the division and the picture changes. Take what you spent on organic work over a year, divide it by the conversions you can attribute to organic in that period, and compare the figure against paid. On a young site the number is often worse than paid; on a mature site with a stock of ranking pages it is usually much better, because last year's pages keep returning while costing nothing further.
What makes the next page cost more than the last one?
You spent the easy terms first. Any sensible roadmap starts where demand meets reachability, which means every page you ship leaves a slightly harder candidate at the top of the list.
Three costs compound as you go down that list. The content itself has to be deeper to compete with established pages. The link and reputation work needed to compete for commercial terms is slower and more expensive than for informational ones. And the stock of pages you already own needs maintaining, so a growing share of your capacity goes to keeping what you have rather than adding to it.
Organic is cheap in aggregate and expensive at the margin. The question is never whether organic is cheaper than paid overall, but whether the next page you commission beats the next dollar in the auction.
What are the signs you have crossed over?
Rising effort for flat return, in a pattern you can see in your own data rather than in a benchmark.
| Signal | What it suggests |
|---|---|
| New pages take longer to rank than they did a year ago | You are working harder terms, or the result layout changed |
| Cost per published page keeps rising | Competing now requires more depth and more review |
| Impressions grow while clicks stay flat | Visibility without clicks; the result is answering for you |
| Most capacity goes to updating old pages | Maintenance is consuming the growth budget |
| Remaining terms are all dominated by larger brands | The reachable list is spent |
Does being summarized in an AI answer change the math?
It changes the denominator. When a result answers the question on the page, an impression can be delivered without a click, so visibility and traffic stop moving together.
Google has shipped AI Overviews into results across many query types, which makes the per-impression value of informational visibility less predictable than it was. That is not an argument against being the source of the answer; being cited is worth something. It is an argument for measuring organic by conversions rather than by sessions, and for weighting pages that need a click to be useful.
Comparing cost per click instead of cost per acquisition. Organic wins the click comparison almost by definition, which is why it is the comparison that hides a channel that stopped paying for itself.
What is worth doing when the next page is not?
Spend on the pages you already have. A page that ranks and converts badly is usually the cheapest conversion you can buy, because the hard part, visibility, is already done.
Three moves are reliably cheaper than a new page. Improve the conversion rate of pages that already get qualified visits. Fix pages ranking just outside the top ten, where a relevance or depth fix is small and the gain is immediate. And strengthen internal linking into the pages that matter, which costs an afternoon and no new content at all.
Should the answer be organic or paid?
Neither, as a standing rule. They buy different things: paid buys immediate, controllable volume at a price that rises with competition, and organic buys an asset that compounds but takes a quarter or more to show up.
What changes over time is the mix, not the choice. A business with no ranking pages and an urgent target should be paying for clicks while it builds. A business with a hundred ranking pages and a rising cost per new page should be harvesting what it has, keeping its Google Search Console data as the record of what still has room, and buying volume only where organic cannot reach.
What number would settle it for you?
Work out what your last ten published pages cost and what they have returned since. If that figure is worse than your paid cost per acquisition and trending the wrong way, your next decision is about harvesting rather than publishing.