Key Takeaways
- Programmatic SEO is a dataset play, not a content hack. The pages that rank carry information unique to each URL. Pages that only swap a city or product name into a shared template are what Google's scaled content abuse policy is built to catch.
- The math changed. Ahrefs found 96.55% of all pages get zero organic traffic from Google, and SparkToro measured 68% of 2026 searches ending with no click. Generating a thousand pages does not beat those odds. Generating a thousand thin pages worsens them.
- Thin programmatic pages are a sitewide risk, not a per-page one. Google's helpfulness signals suppress the whole domain when a large page set scores below the quality threshold, so a bad build can subtract traffic from pages that were already ranking.
- The safe use cases are commercial, comparison, and local queries backed by proprietary data. The unsafe ones are informational 'what is X' pages, which AI Overviews now answer without a click.
- For most founder-run B2B companies, the honest answer is a smaller set of commercial-intent pages tied to a booked conversation, not a thousand URLs you will have to maintain.
The short answer: it depends on your dataset
Programmatic SEO is the practice of generating many pages from one template and a structured dataset, where each page targets a specific long-tail query that would be impractical to write by hand. Zapier built more than 70,000 app-integration pages this way. Preply built roughly 48,000 tutor pages. The technique still works in 2026.
The catch is what makes it work. The pages that rank carry data that is genuinely unique to each URL. The pages that get penalized swap a city name or a product name into an otherwise identical template and call it a page. Google spent the last two years teaching its systems to tell those two apart, and it got good at it. So the real question is not "can I generate a thousand pages," it is "do I have a dataset worth a thousand pages, and does each page answer a question no other page on my site already answers." For most founder-run B2B companies, the answer to that is no, and shipping the pages anyway does more harm than doing nothing.
The math changed before you generated a single page
The pitch for programmatic SEO rests on volume: publish enough pages and some fraction will rank, and the fraction adds up. That logic was always shakier than it sounded, and in 2026 it is weaker still.
Start with the base rate. Ahrefs studied its index of billions of pages and found that 96.55% of all pages get zero organic search traffic from Google, with only 1.94% earning between one and ten monthly visits. That is not a statement about bad pages. It is the default outcome for published content. A programmatic build does not exempt you from it; it multiplies your exposure to it.
Then add what has happened to the clicks that remain. SparkToro's analysis of Similarweb clickstream data found that 68% of US Google searches ended without any click in early 2026, up from 60% two years earlier. AI Overviews now appear on more than 20% of searches, and when they show up they cut click-through to the top result by close to 60%. Most programmatic pages target informational queries, which is exactly the intent AI Overviews answer on the results page. You can rank first and still get almost no visit.
There is a second problem hiding inside the volume pitch. The long-tail queries programmatic pages chase are individually tiny. Ahrefs found that 93% of keywords in its US database get fewer than 10 searches a month. That is the whole reason the tail is worth chasing at scale, but it also means each page is fighting for a handful of visits. If the page does not convert the few people who land on it, or if an AI Overview answers the query before they click, the page returns nothing while still counting against your site's quality score. The upside per page is small and the downside is shared across the domain.
None of this kills programmatic SEO. It changes the break-even point. When most pages earn nothing and a growing share of searches send no click at all, the per-page quality bar has to be much higher for the volume math to clear. Generating more pages does not beat those odds. Generating more thin pages worsens them.
The line Google actually draws
In March 2024 Google added scaled content abuse to its spam policies, and it has enforced the rule through core and spam updates ever since. The official definition is worth reading closely, because it is narrower and more precise than the panic around it suggests: the policy targets "creating many pages where the content adds little or no value" for the primary purpose of manipulating rankings.
Two words carry the weight. The first is "primary." Google has stated repeatedly that how a page is made, by a human, a template, or a model, is not the trigger. A page built by automation and filled with verified, useful data is fine. A page built by hand that exists only to catch a keyword is not. The second is "value." The test Google's systems apply is whether each page answers a distinct query with something a user cannot get from the other pages in the set. A "plumber in [city]" template that produces a thousand pages with no real plumber data, no pricing, and no local context fails that test regardless of how much effort went into the template.
The part that catches operators off guard is the blast radius. Google's helpfulness signals run continuously, not only during named updates, and they apply at the site level. When a large set of pages scores below the quality threshold, the demotion is not confined to those pages. It lowers the ranking ceiling for the whole domain. A programmatic build done badly does not just fail to add traffic. It can pull down the pages that were already ranking before you started. That is the asymmetry worth sitting with: the downside is not a wasted afternoon, it is a net loss.
When programmatic SEO is the right move
The technique earns its place when three conditions hold at once, and each one is a real constraint rather than a box to check.
You have a proprietary or hard-to-replicate dataset. This is the whole game. If the data behind your template is the same publicly available information everyone else can pull, you have no defensible page. The sites that win, like Zapier's integration pages or a comparison site with live pricing feeds, are sitting on inventory or data that no competitor can copy cheaply. Original survey data, first-party usage numbers, real reviews, and licensed datasets all qualify. A rephrased manufacturer description does not.
The queries carry commercial, comparison, or local intent. These are the searches AI Overviews touch least, because the searcher wants to compare options or reach a specific provider, not read a summary. A B2B battery retailer profiled by SEO Tailor grew clicks 200% year over year by aiming its programmatic pages at commercial "car model plus battery" searches rather than informational ones. Deprioritize the "what is [term]" pages. AI answers own that intent now, and winning a citation there takes far more quality than most programmatic builds produce, a point worth understanding before you commit, and one we cover in how to rank in AI search.
You can support and maintain the set. Programmatic pages need a pillar page above them, dense internal links between them, and a pruning rule that removes pages earning nothing after 90 days. This is the piece founder-run teams underestimate. A set that is shipped and abandoned decays under Google's continuous evaluation. If no one owns the maintenance, the build has a shelf life measured in months. The structural work here is the same discipline covered in our B2B SEO strategy guide: a real information architecture that ties the pages into a structure a crawler and a buyer can both follow.
When it is the wrong move, and what to do instead
If you do not have a distinctive dataset, if your target queries are informational, or if no one on the team can maintain the set, programmatic SEO is not your next move. It is a way to convert a quiet afternoon into a sitewide ranking problem.
The alternative is less exciting and works better for the size of company we usually see. Rather than a thousand templated pages, build a small set of commercial-intent pages that each earn their place: a genuine comparison of the tools your buyers actually weigh, a service page for each problem you solve, a page for each segment you serve, written well enough that you would publish it on its own merits. Ten pages a buyer trusts will out-earn a thousand a buyer never sees, and they will not put the rest of your site at risk.
The measure that matters is not pages published or even traffic. It is whether a ranked page leads to a booked sales conversation. That is the number the Demand Engine is built around: search and AI-search visibility that turns into commercial-intent visits, and visits that turn into calls. A page that ranks and converts nobody is a vanity metric with a maintenance cost. Before you generate anything at scale, it is worth running a plain SEO and AEO audit to see whether your existing pages are already competing with each other, and worth checking whether the structured-data work you are counting on actually helps, which we tested in schema markup for AI.
Programmatic SEO is not dead, and it is not a shortcut. It is a scale multiplier that rewards companies with a real dataset and punishes companies without one. Know which you are before you build. If the honest answer is that you do not have the data or the capacity yet, the better spend is a handful of pages that turn search demand into conversations, wired into a system that reports on the one number that pays the bills.
