Programmatic SEO is not dead: lazy AI templates are
Programmatic SEO can still work after AI content saturation. Use real data, proof and page filters before you scale.
Programmatic SEO used to sound like a money printer.
Now it can look like a spam factory with a nicer spreadsheet.
That does not mean programmatic SEO is dead. It means founders no longer get rewarded for making 2,000 thin pages with swapped city names, scraped descriptions and AI filler. Good. The internet has enough beige mush. A bootstrapped founder should not add more.
TL;DR: Programmatic SEO means creating many search pages from a repeatable template and a real data source. It can still work after AI content saturation, but only when each page deserves to exist. Use it for searchable patterns where the data changes the answer, such as locations, product variants, comparison sets, calculators, directories and use cases. Before you scale, define the source data, proof, internal links, crawl rules, page filters and refresh rhythm. If the page would be useless without the template variables, do not publish it.
I am Violetta Bonenkamp, also known as Mean CEO. Through Mean CEO, CADChain and F/MS Startup Game, I have built under constraints long enough to know that speed is useful only when it moves you closer to customers. Fast spam is still spam. It just reaches the wall earlier.
The only defensible programmatic SEO now starts with proof. E-E-A-T and first-party data gives that proof a stronger source layer. If every page is a template with no data, no source, no real difference and no buyer use, AI has made it cheaper to produce and easier to ignore.
What Programmatic SEO Means
Programmatic SEO is the creation of many search-targeted pages from a structured data source and a reusable page pattern.
That sounds technical. The idea is old.
Think of:
- City pages for a service.
- Product variant pages.
- "Best X for Y" pages.
- Directory pages.
- Comparison pages.
- Calculator pages.
- Template pages.
- Tool-connection pages.
- Local pages.
- Industry use-case pages.
Ahrefs’ programmatic SEO guide defines the tactic as creating keyword-targeted pages in an automatic or near automatic way, usually from data such as product prices, weather or location information. It also points to examples from Zapier, Zillow, G2, Webflow, TripAdvisor, Yelp and Amazon.
So no, the method is not new.
What changed is the cost of producing weak pages.
AI made thin pages cheap. Search systems, answer engines and users now have less patience for pages that exist only because a spreadsheet can multiply.
Why AI Content Saturation Changed The Game
The web is filling with AI-assisted content.
Ahrefs analyzed 900,000 newly created pages in April 2025 and found that 74.2 percent contained AI-generated content. Pure AI pages were a smaller group, but mixed human and AI content was everywhere.
Then look at the ranking side.
Graphite’s research on AI content in search and answer engines found that 86 percent of articles ranking in Google Search were written by humans, and only 14 percent were generated using AI. It also found that 82 percent of articles cited by ChatGPT and Perplexity were human-written, while 18 percent were AI-generated.
The lesson is not "never use AI."
The lesson is that cheap page production is no longer the advantage. The advantage is data, filtering, judgment and proof.
Google’s generative AI content guidance says AI can help with research and structure, but using AI or similar tools to create many pages without added worth for users may violate the scaled content abuse policy. Google’s spam policies also explain that sites violating spam policies may rank lower or not appear in results.
That should make lazy page factories nervous.
It should make serious founders calmer.
If you have real data and useful filters, programmatic SEO is not the problem. Empty scale is the problem.
The Founder Test Before You Create 1,000 Pages
Before you build a programmatic SEO project, ask one brutal question:
Would this page help the buyer if Google never existed?
If the answer is no, stop.
Then ask:
- Does each page answer a distinct search intent?
- Does the page use real data that changes by row?
- Does the page have a source that can be checked?
- Does the page help a buyer compare, decide, filter or act?
- Does the page link to the next useful page?
- Does the page have a reason to be refreshed?
- Does the page avoid doorway behavior?
- Does the page deserve indexation?
Doorway-style pages are dangerous because they create many near-identical pages to capture search demand while offering little help. Programmatic SEO becomes defensible when the page set makes the site more useful instead of larger for its own sake.
Programmatic pages need clean facts. Use structured data SEO for AI retrieval to make facts, entities, prices, and proof easier for machines to read. Machines need to understand what the page is, what data changed, which entity is being described and where the proof sits.
The Programmatic SEO Quality Gate
Use this table before building a page set.
Location, service scope, price range, proof and local constraints
Each city has real demand and one detail that changes the decision
Product specs, use cases, limits, pricing and availability
The variant changes the buyer answer beyond the URL
Feature facts, price facts, buyer fit and tradeoffs
You can compare honestly without inventing weak differences
Verified entries, categories, filters and update dates
The directory saves the buyer time and stays maintained
Formula, inputs, assumptions and result logic
The result changes based on user input and helps a decision
Use case, preview, instructions and source file
The page helps the reader copy, adapt or use the asset
Definition, examples, related entities and source links
The term needs disambiguation and links into a larger topic cluster
Buyer segment, workflow, problem, proof and next action
The use case exists in sales calls or customer research
This table is your first kill switch.
If you cannot fill the data column, do not publish the page set.
Good Programmatic SEO Starts With A Dataset
The page template is not the hard part.
The dataset is the business.
A useful dataset can come from:
- Product catalog fields.
- Locations served.
- Pricing rules.
- Public grant databases.
- Customer objections.
- Search Console queries.
- Internal search terms.
- Tool pairings.
- Use-case tags.
- Event calendars.
- Public benchmarks.
- Original research.
- Customer support themes.
- Survey answers.
For bootstrapped founders, the best dataset often comes from sales and customer work, not some expensive tool. If ten prospects ask the same question in slightly different ways, you may have a page pattern. If each question requires a different answer, you may have a manual content cluster instead.
The F/MS article on semantic SEO for female entrepreneurs makes a useful point for small teams: founders who build from customer feedback and lived work can create topical depth that bigger competitors cannot copy easily. That is exactly what programmatic SEO needs now. It needs source material that belongs to the company.
Bad Programmatic SEO Starts With A Keyword Sheet
A keyword sheet is not a product.
It is a list of temptations.
Bad programmatic SEO starts like this:
- Export thousands of keyword variations.
- Build one generic template.
- Ask AI to fill the blanks.
- Publish every combination.
- Hope Google cannot tell.
This is how founders confuse automation with distribution.
Good programmatic SEO starts differently:
- Pick one repeatable buyer job.
- Build one excellent page manually.
- Prove that the page helps a real buyer.
- Identify which data fields make the answer change.
- Create the page template only after the manual page works.
- Publish a small page set.
- Watch indexation, clicks, conversions, branded search and assisted sales.
- Remove or merge weak pages quickly.
Bootstrapped founders need repeatable systems. The F/MS guide on creating useful content without hiring a marketing team can help a founder turn daily learning into useful content without hiring a department. But repeatable does not mean careless. It means one good answer can become a small, maintained library when the data deserves it.
Crawl Waste Can Kill The Whole Project
Programmatic SEO can create a crawl mess.
That is not a small technical footnote. It can make the site slower to discover and harder to trust.
Google’s faceted navigation crawling guide warns that URL parameters can create very large URL spaces, causing overcrawling and slower discovery of useful URLs. It also says crawling faceted URLs can cost sites large amounts of computing resources.
Plain founder translation:
If your filters create endless URLs, bots may spend time on junk.
So before publishing a page set, decide:
- Which pages should be indexed.
- Which filtered URLs should stay blocked.
- Which pages need canonicals.
- Which parameters should not create crawlable URLs.
- Which thin pages should never ship.
- Which old pages should be merged or removed.
This is where programmatic SEO meets cash discipline. Every bad URL consumes time, crawl attention, server resources and sometimes developer time. A bootstrapped founder should be allergic to that kind of silent waste.
Build The Minimum Useful Page Set
Do not start with 10,000 pages.
Start with 20 to 50.
Pick a page type with clear business logic:
- A location where people actually search.
- A comparison buyers ask about.
- A tool pairing that creates real demand.
- A directory slice that saves time.
- A calculator that answers a money question.
- A use case that appears in sales calls.
Then give every page:
- A unique H1.
- A unique answer above the fold.
- Data fields that change the page meaning.
- A source or proof point.
- A clear next step.
- Links to parent and sibling pages.
- A date or refresh note when data changes.
- A reason to be indexed.
Product-led pages should come from actual product or data use. Product-led SEO for tiny SaaS teams is the stricter version of that test: if the product cannot create useful pages, the programmatic layer may just be decoration.
Use AI For The Boring Parts, Not The Proof
AI can help a programmatic SEO workflow.
Use it for:
- Drafting page patterns.
- Grouping search queries.
- Cleaning data fields.
- Suggesting internal link targets.
- Creating QA checklists.
- Finding duplicate wording.
- Turning customer phrases into clearer headings.
- Summarizing public sources.
Do not use AI as the witness.
AI cannot prove your price is current. It cannot know that a city page reflects local delivery. It cannot know that a directory entry is still alive. It cannot know that a tool pairing works unless your product, data or team proves it.
This is the line founders should respect:
AI can help you package evidence. It cannot replace evidence.
Programmatic SEO becomes much harder to copy when the dataset is original. Use original research as a backlink engine to make data, citations, and backlinks support the same proof engine. If the data is public, your angle and freshness must be better. If the data is yours, you have a stronger moat.
The Page Filter That Saves Founders From Spam
Before any generated page goes live, run a filter.
Do not publish the page if:
- The page has fewer than three unique facts.
- The answer is the same as the parent page.
- The keyword has no buyer intent.
- The data is stale.
- The location or category is fake.
- The page has no internal links.
- The page has no next step.
- The source cannot be checked.
- The page exists only to catch a keyword variation.
- The page would embarrass you if a buyer opened it.
That last test is primitive and useful.
If you would not send the page to a prospect, why are you sending it to Google?
Refreshes Matter More Than Publishing Bursts
Programmatic SEO is not done when the pages go live.
That is when the maintenance bill starts.
Track:
- Which pages get indexed.
- Which pages get impressions but no clicks.
- Which pages get clicks but no business signal.
- Which pages duplicate intent.
- Which pages cite stale facts.
- Which pages have weak internal links.
- Which pages answer a query that AI answers fully.
- Which pages deserve merging.
- Which pages deserve deletion.
Small sites often get more from improving proven pages than publishing more pages into the void. Use content refreshes as the cheapest SEO moat to improve pages that already have proof before publishing more thin pages.
A healthy programmatic SEO system has a deletion habit.
That sounds harsh.
Good. Pages should earn their place.
What To Do This Week
Use the small founder version.
Pick one programmatic idea and write it on one page:
- Page pattern: What repeatable page could exist?
- Buyer job: What decision does it help?
- Dataset: What real data changes by page?
- Proof: What source or company evidence supports it?
- Filter: Which pages must never publish?
- Internal links: Which parent, sibling and proof pages does each page need?
- Refresh rule: What gets checked monthly?
- Business signal: What action should the page create?
Then build one manual example.
Do not automate yet.
Send the manual page to a prospect, customer or founder friend who fits the buyer. If she says, "This is useful," you may have a pattern. If she says, "This feels generic," you have a warning.
Programmatic SEO should multiply proof.
It should not multiply wishful thinking.
FAQ
What is programmatic SEO?
Programmatic SEO is the creation of many search pages from a repeatable template and structured data. The page pattern stays similar, but the data changes the answer. Good examples include location pages, comparison pages, product variant pages, directories, calculators and use-case pages. The method works only when every page helps a real searcher.
Is programmatic SEO dead after AI content saturation?
No. Lazy programmatic SEO is dying. Programmatic SEO still works when pages use real data, clear intent, useful filters, internal links and proof. AI content saturation makes generic templates easier to produce and less interesting to search systems. Founders need fewer pages with stronger evidence, not more pages with thinner copy.
Can a bootstrapped founder use programmatic SEO?
Yes, but the page set should start small. A founder can use customer questions, product data, pricing rules, locations, public datasets or original research to create a tight page pattern. Start with 20 to 50 pages, measure indexation and business signals, and expand only when the pages prove useful.
What makes a programmatic SEO page worth publishing?
A page is worth publishing when it answers a distinct buyer question, uses data that changes the answer, includes source material, links to relevant pages and gives the reader a clear next step. If the page is just a template with swapped words, it should not go live.
How is programmatic SEO different from spam?
Programmatic SEO helps users navigate a large set of real options, facts or use cases. Spam creates pages mainly to catch search traffic. The difference is intent and usefulness. If the page saves time, compares honestly, uses real data and would help a buyer even without Google, it is much safer.
What data can I use for programmatic SEO?
Use data you can maintain. That may include product fields, locations, prices, availability, customer objections, public benchmarks, directory entries, comparison facts, templates, calculators, search queries, support themes or original research. The dataset should make each page different in a way that matters to the buyer.
Should I use AI to write programmatic SEO pages?
Use AI for structure, pattern drafting, query grouping, data cleanup and review checks. Do not use AI to invent facts, proof, prices or comparisons. AI can package evidence, but the evidence must come from your product, customers, research or sources.
How many programmatic SEO pages should I publish first?
Publish a small test set first, usually 20 to 50 pages. That is enough to see whether Google crawls them, whether they get impressions, whether users click and whether any business signal appears. Publishing thousands of pages before the pattern works is founder ego wearing a spreadsheet.
What are the biggest programmatic SEO mistakes?
The biggest mistakes are publishing pages with no unique facts, creating endless URL combinations, copying public data without a better angle, ignoring internal links, failing to refresh pages and letting AI fill pages with generic filler. The quiet killer is crawl waste: too many weak URLs can distract bots from useful pages.
How do I measure whether programmatic SEO is working?
Track indexation, impressions, clicks, conversions, branded search, assisted sales conversations and refresh needs. Also track deletion candidates. A programmatic SEO system should produce useful pages and remove weak ones. If the page set grows but sales conversations do not improve, the project may be search vanity.
The Bottom Line
Programmatic SEO is not dead.
The lazy version is having a very public identity crisis.
AI made it cheap to produce thousands of pages. That is not the same as earning trust, citations, rankings or customers. A founder who wants programmatic SEO to work now needs a real dataset, a useful page pattern, crawl discipline, internal links, proof and a deletion habit.
Start with one manual page.
Make it useful.
Then multiply only the part that deserves to be multiplied.