Content

Good vs. Bad Programmatic SEO: Lessons From a Decade of Scalable Content

Search Engine Journal published 'Good Vs. Bad pSEO: Learnings From A Decade Of Building Scalable Content Engines' on November 10, 2026, and it draws a line the industry has been blurring for years. The author has been building programmatic SEO systems since 2016, from static templates to natural language generation engines such as Rosae NLG to AI-assisted pipelines. His conclusion is that most of what passes for programmatic SEO today is not programmatic SEO at all — it is mass AI publishing wearing the label.

A publishing method, not a content type

The piece starts with a definitional correction: programmatic SEO is a publishing method, not a content type. It is the practice of generating many search-optimised pages at once by combining a single template with a structured dataset, so each page targets a specific, repeatable search query. The word 'programmatic' does not appear in Google's spam policies at all — what the policies target is scaled content abuse, the mass production of low-value pages, whether a human, an automation system, or a combination of the two produced them.

That distinction is the key to the whole article. Google does not care how a page was made; it cares whether the page is useful. A page built on real data and drafted with AI can still perform well, while a hand-written page with no real data behind it has no more claim to rank. The method is neutral; the output is what gets judged.

Six checks that tell the two apart

The article offers six checks that look underneath the output, because good and bad programmatic SEO look identical from the outside. The checks are a diagnostic: if one row fails, you have a workflow problem to fix; if several fail, you are mass publishing, whatever you call it. In substance they cover real search demand behind each page, unique and verifiable data on every URL, pages that help a visitor finish a real task, logical internal links, a controlled rollout measured before expansion, and indexing outcomes that prove the pages earned their place.

None of the checks asks whether AI wrote the page — that question is a distraction. What matters is whether each generated page carries something a visitor can use — a dataset, a fact, a comparison, a tool — that no other page has in quite the same form. Mass AI publishing fails because it starts with the template and works backwards to the data, producing thousands of pages whose only difference is a swapped keyword.

Google's policy backdrop

Google has been clear about where it stands. On March 5, 2024, it strengthened its scaled content abuse policy to apply whether automation, humans, or a combination produces the pages, and later reported 45 percent less low-quality, unoriginal content in search results. That is the policy context in which every programmatic SEO programme now operates: scale is not a defence, and human involvement is not a defence either. Only value is a defence.

Enforcement has continued into 2026. On September 7, 2026, John Mueller replied on Bluesky to a site that generated pages from every combination of domain names, technologies and attributes, writing that programmatic SEO of that kind often leads to a site that is spam, borderline spam, or low quality — and that pages already published may keep shaping how Google judges the whole site. A bad batch does not just fail on its own; it stains the domain.

The decade's arc, and the test that matters

The author traces his own arc from 2016 static templates through natural language generation to today's AI-assisted pipelines, and notes that the pages he built in 2016 and the ones he builds today look very similar. What changed is the cost of producing them — which is exactly why the abuse has exploded. When ten thousand pages cost almost nothing, the temptation to generate them regardless of demand becomes overwhelming. He has presented the work at conferences, most recently on scaling pSEO at the Shenzhen SEO Conference.

The piece also cites two tests. Eli Schwartz argues that spinning up pages just because you can is not SEO but programmatic spam, and that creating pages because they benefit many users is good SEO, whether you create them programmatically or manually. Danny Sullivan, then Google's Search Liaison, shared an intent test at a 2025 event: was the intent 'I'm going to get that traffic' rather than 'some user actually expected it' if they came to your site directly? Both tests point the same way — start from the user, not from the traffic.

Why it matters for SEOs and link builders

For SEOs, the article is a reality check at a moment when AI makes it cheap to publish at a scale unthinkable five years ago. The economics have flipped: the constraint is no longer production, it is judgement. The teams that win with programmatic SEO in 2026 will have a proprietary dataset, a real reason for each page to exist, and the discipline to roll out in small batches and measure before expanding. Everyone else is building a bigger target for the next spam update.

For link builders, the lesson is about which pages deserve links. A programme of thin pages is not a set of link opportunities — it is a set of liabilities, and linking to them transfers risk, not authority. The pages worth building links to are the ones that pass the article's checks: unique data, real demand, a task completed. That is a smaller set than most programmes produce, and that is precisely the point.

Sources

Sources: Search Engine Journal, November 10, 2026.

Related reading on this site: Website Migrations Can Hurt SEO Even When URLs Don't Change, 3 in 4 Homepages Lost Organic Clicks in an Analysis of 110 Sites, SEO Split Testing in 2027: How to Prove a Change Actually Worked, The New Organic Growth Stack: Search, Social, Communities and AI, Ziff Davis Says Organic Google Referrals Fell 37% as AI Citations Hit 2.7 Billion, and Tripadvisor Says SEO Will Still Weigh on Growth in 2027.

More from Linkbaits

Turn this playbook into links

Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.

Start free trial