Programmatic SEO, Explained for Beginners

Marketing reports, apps, and automations, handled. Meet Letaido.

Get Started

Marketing reports, apps, and automations, handled. Meet Letaido.

Get Started

The only AI marketing platform built on Ahrefs’ proprietary web index.

Get Started

The only AI marketing platform built on Ahrefs’ proprietary web index.

Get Started

“The AI marketing platform we needed didn’t exist. So we built it.”

Tim Soulo

Tim Soulo, CMO at Ahrefs

Get Started

“The AI marketing platform we needed didn’t exist. So we built it.”

Tim Soulo

Tim Soulo, CMO at Ahrefs

Get Started

Content Marketing

General SEO

Portrait of Louise Linehan

By Louise Linehan

Content Marketer at Ahrefs

Updated: September 2, 202615 min read

322 Organic traffic243 Linking websitesArticle performance data from Ahrefs

Contents

Programmatic SEO refers to the creation of keyword-targeted pages in an automatic (or near automatic) way.

It provides a way for companies to create thousands of website pages targeted at thousands of keywords—without having to design, write, and publish pages manually.

Companies like Canva and Notion use programmatic SEO to generate millions of pageviews each year.

Programmatic pages are built from data like product prices, weather, or location information, with one page for every row in the dataset.

Creating them at scale used to be a job for web developers.

Now an AI assistant connected to SEO data can do most of it, and the rest is judgement: picking the right keywords, the right data, and the right things to say.

We'll show you how.

Using Ahrefs Site Explorer data, we’re able to see a site’s programmatic content growth and organic traffic growth side-by-side.

Here are some sites that seem to be getting it right.

NutriScan’s multilingual nutrition directory

URL: <https://nutriscan.app/calories-nutrition/> Estimated pages: ~4K Estimated monthly organic traffic: ~182K

NutriScan is an AI food scanner and calorie-tracking app.

Its four main language folders grew from roughly 100 ranking pages in October 2025 to ~4K in July 2026.

In all, it attracts ~182K monthly organic visits by targeting the long tail of food-name searches combined with terms such as “calories,” “protein,” “nutrition,” and “weight loss.”

Line graph: NutriScan's organic traffic grew from 0 to 190K, and pages ranked from 5 to 4.5K, from Jan

It has programmatically created thousands of pages for individual foods, restaurant menu items, and branded products, then reproduced the directory in Spanish, French, and German.

An example programmatic page for a McChicken, showing the standardized title, quick facts, health-goal analysis, blood-sugar guidance, comparisons, FAQs, and related foods.

From a moussaka to a McDonald’s McChicken, every page follows a similar kind of template, featuring:

  • Calories and macronutrients
  • Health benefits
  • Myths
  • Substitutions
  • FAQs

Ad Hoc News’ stock-news directory

URL: <https://www.ad-hoc-news.de/boerse/news/> Estimated pages: ~72K Estimated monthly organic traffic: ~1.74M

Ad Hoc News is a German news site that publishes stock-market stories at scale under /boerse/news/.

The directory expanded from about 14K ranking pages in August 2025 to 72K in July 2026 and now attracts an estimated 1.74 million monthly organic visits.

Ahrefs data showing the rapid expansion of Ad Hoc News’ ranking pages and estimated organic traffic.

Pages cover individual companies, share-price movements, earnings, analyst commentary, and other market events.

Ad Hoc News’ stock-news directory, showing the repeated stream of company- and event-led articles.

Each headline combines a company or ticker with a timely event. Think: stock rising, falling, reaching a record, or reacting to earnings.

For example, here's a story about Meta’s stock and AI spending and another on Nvidia’s share price and AI demand.

Canva’s feature pages

URL: <https://www.canva.com/features/> Estimated pages: ~310 currently ranking Estimated monthly organic traffic: ~13M

Canva creates feature landing pages targeting searches for specific design tools and capabilities, like photo editors, video editors, PDF converters, and AI-powered tools.

The directory grew from just 23 ranking pages in early 2022 to ~310 today. Organic traffic grew even faster, rising from ~20K monthly visits to roughly 13 million over the same period.

A line graph showing Canva's organic traffic (blue) growing from 20K to ~12M, and ranked pages (orange) from 23 to ~310.

These pages have a repeatable architecture that Canva can generate across hundreds of different tools and features.

For example, its background remover page follows the same basic structure as other feature pages, with a feature-specific headline, benefits, examples, FAQs, and calls to action that take visitors directly into the tool.

A Canva webpage for “Background Remover” with an example image of a person on a transparent background, an upload area, and instructions.

The copy seems to be more bespoke than classic database-driven programmatic content, and is likely individually edited, even if AI assists with its creation.

But the repeatable page architecture means Canva can systematically target hundreds of feature-specific searches without building every landing page from scratch.

Notion’s template category directory

URL: <https://www.notion.com/templates/category> Estimated pages: ~600 currently ranking Estimated monthly organic traffic: ~204K

Notion is a productivity and collaboration platform.

In late 2024, it launched hundreds of template category pages, growing from almost no ranking pages to more than 500 in just a few months.

Organic traffic grew just as quickly, rising from almost nothing to ~140K monthly visits by early 2025, and eventually reaching ~204K today.

Line chart showing Notion's template category pages (orange, right axis) and organic traffic (blue, left axis) from Aug 2024 to Jul 2026. Organic traffic grew to

These pages are programmatically created around Notion’s template inventory, targeting use cases like product management, marketing, personal finance, and website building.

Rather than manually writing a landing page for every use case, Notion’s content scales with its template catalogue and category taxonomy.

Notion's template marketplace showing “Categories.” A grid of “Work templates” includes Product, Marketing, Design, Engineering, Startup, AI, Operations, and HR categories, each with an icon, template count, and small preview.

QuillBot’s AI writing tools library

URL: <https://quillbot.com/ai-writing-tools/> Estimated pages: ~130 Estimated monthly organic traffic: ~464K

QuillBot has built out a library of AI writing tools, with dedicated pages targeting specific writing tasks and use cases.

The library went from virtually no search presence in early 2025 to ~464K monthly visits in August 2026.

Line graph showing QuillBot's organic traffic growing from 0 to 464K and ranked pages

What's interesting is that the traffic growth didn't require constant page additions. Once the initial build was done, the pages (~130 of them) kept earning more traffic from the same URL set.

A website showcasing Quillbot's AI Writing Tools. Features include text generation, grammar checker, and content creation tools.

Every page targets a user need (write an email, generate a title, create a cover letter) and sits within QuillBot's tool ecosystem, with links to complementary tools and optional paid upgrades.

Before you hit publish on thousands of pages, it's worth considering the words of Google's John Mueller: “Programmatic SEO is often a fancy banner for spam.”

Any company that publishes a huge number of very similar pages runs the risk of creating thin content: content that offers little to no value to the end user.

Like any other page, programmatic content needs to satisfy user intent (and not violate Google’s spam policies).

Head to LinkedIn and you'll see many examples of sites that have tanked in Google for this very reason.

Lars Lofgren has spent months begging people to stop doing programmatic SEO, after seeing company after company whose rankings got demolished once they scaled into the thousands of pages.

A LinkedIn post by Lars Lofgren, Fractional VP of Marketing, warns against programmatic SEO, calling it risky and a cause of ranking drops.

And Lily Ray thinks certain programmatic page types become a liability rather than an asset once a site scales far enough, and has tracked over 70 companies caught in Google's early 2026 updates after scaling content.

Not everyone agrees programmatic content is worth avoiding, though.

Zak Perez has a client who scaled AI content across 10,500 pages, and saw consistent, across-the-board ranking gains over the course of years and throughout multiple Google updates.

LinkedIn post by Zak Perez on AI-generated content performing well over two years, with Joshua Squires' comment and Perez's reply.

And our own Director of Content Marketing, Ryan Law, makes a related case for scaling something other than articles: Ahrefs' free tools are consistently among our highest-traffic pages, holding steady even as AI answers eat into clicks elsewhere.

Searchers want to do something, not read about it, and a calculator is a lot harder for Google to summarize out of existence than a paragraph is.

Marketers keep declaring tactics like pSEO (programmatic SEO) worthless the moment they get more complicated, which cedes the opportunity to whoever sticks around.

According to our own research, publishing at scale is only a risk if you don't have anything worth publishing at scale.

Further reading


Google Doesn’t Punish AI Content; It Punishes Bad Content (331k Pages Studied)


Not sure whether to invest in programmatic SEO? We created a 5 question quiz so you can check your own project before building anything.

Should you invest in programmatic SEO?

To create programmatic SEO content, there are five main steps: find a group of similar keywords, decide on content for your template, collect useful data, use AI to turn it into pages, then automate that whole process.

Below, I’ll walk through each step on a project I run myself: templated articles showing the “fastest growing X companies” in different industries, built using Ahrefs’ organic traffic and search volume data.

Programmatic content targets lots of similar keywords with one page template, so you need a phrase where only one part changes.

For example, “Fastest growing [SaaS] companies”, “Fastest growing [AI] companies”, “Fastest growing [tech] companies”.

Connect your AI assistant or agent to an SEO MCP and first ask it to pull keywords around the topics you want to write about. Then, within that set, ask it to find and group any keywords that share a repeating phrase (as above).

Starter prompt


Using the Ahrefs MCP, pull matching terms and questions for these seed keywords: [seeds]. Group them by the repeating part of the phrase. For each group give me the number of variations, combined monthly volume, median keyword difficulty and the median DR of the sites currently ranking. Ignore groups with fewer than 50 variations, and rank what’s left by combined volume.

Here’s what came back for me.

Chat with Claude connected to the Ahrefs MCP. The prompt asks it to pull matching terms for “fastest growing companies” in the US and group them by the repeating part of the phrase. The reply identifies the pattern “fastest growing [category] companies” with 169 variations, 8,330 combined monthly US searches across 122 distinct categories and a median keyword difficulty of 9, followed by a table of example categories including AI, cybersecurity, tech, ecommerce, edtech, software, fintech, healthcare and SaaS companies with their search volume and difficulty.

169 variations across 122 categories, and a median keyword difficulty of 9.

Each of these queries wants the same thing: a ranked list of companies in one category, where only the category changes—which makes it a good candidate for pSEO.

You can also find these keywords manually in Ahrefs.

Run Matching Terms in Keywords Explorer, set the Include filter to the repeating phrase and a minimum volume of 10, then read the count and volume off the top of the report.

Ahrefs Keywords Explorer showing Matching terms for “fastest growing companies” with an Include filter on “fastest growing” and “companies” and a minimum volume of 10, returning 356 keywords with 14K combined US volume. The table below lists variations such as fastest growing ai companies, fastest growing private home health companies us and inc 5000 fastest growing companies, each with its category, keyword difficulty and monthly search volume.

That repeatable template is what you’re looking for.

Every page in a programmatic SEO set needs the same furniture, so look at what the pages currently ranking all have in common.

Ask your AI to use the same SEO MCP to fetch the top ranking pages for your keywords and report the sections they share.

Starter prompt


For these 10 keywords, fetch the top 10 ranking pages each. Tell me which sections, data points and page elements appear on most of them, and which appear on only one or two. List the ones I’d need to include to compete.

Here's what I saw after entering that prompt...

A screenshot of a competitive content analysis outlining common features on ranking pages for “fastest growing companies.”

Ask which sections show up repeatedly, infrequently, and which ones represent genuine gaps.

For my article set, the must-haves included:

  • A jump-link table of contents
  • A stated methodology
  • An explicit “last updated” date (refreshed monthly)
  • A per-company block carrying the same six data points every time
  • The growth figure and company name in the H2
  • A trend chart
  • A short write-up including

The Supabase entry from the Ahrefs fastest growing companies post: a heading reading “1. Supabase (+370%)”, a brand search volume chart with a forecast, then six data points listing brand search growth, forecast, monthly brand searches, organic traffic, year founded and location, followed by a “What they do” write-up.

You can also do this but manually using Ahrefs.

Open SERP Overview in Keyword Explorer, and hit “Identify intents” ¹.

This will give you an overview of What readers expect to see when they hit the SERP for your keyword.

The Ahrefs SERP overview for “fastest growing ai companies” with two areas marked. The first, on the right, is the Identify intents panel, which groups the results into intents and gives each a share: emerging AI startups to watch at 44%, AI companies hiring and scaling at 34%, recently funded growing AI companies at 22%, and fastest-growing AI company lists at under 1%. The second, in the middle, is the Page type column, where the top results are tagged as listing collections and listicles rather than articles.

Check out “Page type”² for more clues. Then open the top results for a handful of keywords and see what they share.

Monitor successful AI generated pSEO content


Letaido reaches extra Ahrefs data beyond what the API has access to, including AI content level.

A screenshot of a content analysis tool, showing an article on Google Search Operators with “Moderate” AI content level and a 55% “Mostly human” ratio.

This lets you check how much of any site’s top pages was written by AI—including the programmatic pages currently outranking you.

That way, you can see for yourself what AI generated pSEO content Google is actually rewarding.

This is what is actually going to make your programmatic SEO content unique and, hopefully, interesting to read.

You have three options: proprietary data that’s yours alone, public data that anyone can license, or scraped data, meaning figures other websites have published that you collect for yourself—though, bear in mind, scraping can raise copyright issues.

Sidenote


If you decide to scrape, Firehose can notify you whenever your source publishes new numbers. That way, the data fueling your pSEO content stays current without you having to constantly check on it.

Whichever option you choose, ask your assistant and SEO MCP about the top ranking pages and their data, so you’re not repeating the same source as everyone else.

Starter prompt


For these keywords, fetch the top 10 ranking pages each and list the data sources they use: named datasets, cited indices, links to data providers, and any proprietary data they claim. Tell me how many pages use each source, and flag anything used by more than half as saturated.

Here's what that brings back...

A document outlining competitive content analysis for fast-growing companies, detailing data sources used by various companies/reports in a table.

Most lists like ours rank companies by funding raised, which is public and heavily reused.

We decide to rank by growth in brand-name searches/organic traffic instead, using Ahrefs’ proprietary data.

Tap into 50+ data sources in Letaido


Letaido already comes with multiple integrated data sources that you can draw on to scale your content.

For instance, if you work in recruitment you could use Firehose to watch job board listings, and create [job title] salaries in [city] content that always reflects the latest data.

The available connectors list in Letaido, showing 2,339 connectors with 2,223 actions and 116 triggers. Firehose, with 11 capabilities, and HubSpot, with 79, are both highlighted, alongside others including GitHub, Gong, Google Ads, Google Search Console and Hunter.

Or, if you work in SaaS, you can pull your reason won fields from HubSpot, so that every one of your “best [tool] alternatives” page says something about why customers actually switch to you.

Now it’s time to turn one row of data into one page. Your template needs to stay fixed, while the data and writing change within it.

Follow these steps:

1. Upload your seed list: Provide a CSV where each row represents a new page section. For my project, this meant mapping out company names, domains, headquarter locations, and founding years.

The seed list behind our fastest growing companies project: a CSV called SaaS_companies.csv with 2,529 rows, one company per row, showing columns for company name, URL, founded year and locality. Visible rows include Supabase, Figma, Notion, Miro, Calendly, Deel, Loom, Ramp, Airtable, Hopin and Manychat. A caption reads that the spreadsheet decides which pages exist and the AI decides everything else.

2. Supply your keywords: In your prompt, include your keyword pattern from step 1.

3. Connect your data sources: Make sure your MCPs and APIs are hooked up, and upload any other relevant data sources.

4. Give direction on the template: Tell the AI:

  • Which parts need to stay fixed: The content that needs to be identical on every page, like the headings and layout.
  • Which parts to take straight from the data: The data that needs to be fetched (e.g. organic traffic data), from which data source (e.g. Ahrefs MCP), how it should be presented (i.e. PNG traffic chart), and where it needs to go (i.e. slotted in per row).
  • Which parts need to be freshly written per page: Prompt the AI on how to draft copy based on the data. For instance, I asked Letaido to cover three main points for each trending company: what the company does, their most recent funding round (with outbound citations), and possible reasons for their growth.

5. Build the workflow and save it as a skill: Before you prompt the AI to build your actual pages, make sure you’ve nailed the workflow. For example, my prompt went something like this: “Read my CSV, look up [the numbers] for every row, save the results to a file, then write one page per row using my keywords and template notes.” At this point, I asked Letaido to build a self-learning skill for me which auto-updates based on my ongoing feedback. I recommend doing this if you can!

6. Test on 10 rows first then run the full list: Read your pages properly and make sure you fix template issues before scaling.

For my lists, Letaido then goes away and pulls the companies, writes each company’s profile based on Ahrefs data, renders a growth chart per profile, and assembles the article.

A generated programmatic page in Letaido for Supabase, page one of twenty-five. It shows brand search growth of 366%, 129,000 monthly brand searches, a rank of one of twenty-five, a line chart of brand searches over twelve months, and a written paragraph describing the growth. A line underneath lists which data the paragraph was written from.

The write-ups all differ based on the underlying data, and Letaido flags any claim it isn’t sure about, so I can check it before publishing.

The free route


If you don’t have access to Letaido, Claude Code, or an AI agent, you can programmatically scale your content for free using just a formula in Google Sheets.

You can make a copy of our example Google Sheets database here.

You need three tabs:

  • Keyword data: Use this to build each page’s title and URL.
  • Data source: Upload your data for each page manually.
  • Formula: Write your page copy, once. Just put your sentence in quotes and drop in a cell reference wherever a value should appear. For example: =“Cost of living in ”&A2&“ is ”&B2&“, ranking ”&C2&“ out of 50 states.”

The formula joins your fixed text with values from your cells.

If you want the sheet to actually write for you, use Google Sheets’ built-in =AI() function on Workspace plans with Gemini, or the GPT for Sheets to use ChatGPT or Claude, so a formula like =GPT(“Write two sentences on why ”&A2&“ is growing, using only these numbers: ”&B2) drafts a different write-up for every row.

A spreadsheet with a formula in the bar at the top joining cell references into a sentence. The rows below list Supabase, Manychat, Airwallex, Hopin and Eldorado.gg with their growth percentage, monthly searches and founded year, and a generated copy column where every row reads identically apart from the numbers.

Then you just drag the formula down.

Every new row generates a new URL, page title and body copy, so you can create 50 pages with the drag of a mouse.

There are only two things left to do now.

1. Connect to your CMS and publish: Connect to your CMS (e.g. WordPress) via Claude Code or Letaido, and instruct the AI to autonomously upload the pages to your CMS.

2. Set up an ongoing automation: Once you’re happy, set up an automation so your data refreshes at your preferred cadence. With Claude Code that means setting up scheduled, API, or GitHub-triggered tasks that run directly on Anthropic’s cloud infrastructure.

Sidenote


You can’t run automations locally in Claude Code, because background tasks stop firing the moment your laptop goes idle.

With Letaido, it’s just a case of ask and you shall receive.

My automation rebuilds every list monthly from fresh data and files a new draft, so the numbers never go stale.

A Letaido monthly rebuild that ran on 1 August, with ticks confirming it pulled fresh data for 25 companies, rebuilt every page and chart, and filed a WordPress draft for review. Below is a list of pages to act on: one with no visits in 90 days marked remove, one down 46% marked improve, and one ranking for nothing in the top 50 marked redirect.

Keep every programmatic page you’ve created inside one folder, such as /salaries/, so you can track the whole project in Ahrefs using Site Explorer’s Path/Prefix mode.

After three months, check:

  • Organic traffic: Is the project growing, holding steady, or fading after an initial spike?
  • Top pages: Which pages still get little or no traffic?
  • Organic keywords: Are individual pages ranking for their intended terms?
  • GSC performance: Are impressions and clicks growing? Is click-through rate falling?

A dashboard showing a performance graph for canva.com/features/. The graph displays “Organic pages” (yellow line) and “Organic traffic” (orange line) from June 201

You can also add representative keywords to Rank Tracker and tag them as “pSEO” to monitor the project separately from the rest of your site.

Ahrefs Rank Tracker screenshot showing the “Tag” filter open, with “pSEO” selected as a rule to filter data.

Review the results monthly.

Improve pages that are losing traffic, redirect pages that overlap, and remove or de-index pages that still earn nothing after a fair testing period.

The goal is to keep the pages that prove useful and stop weak pages from quietly piling up.

Final thoughts

Creating programmatic SEO content is no mean feat, but it’s a lot easier lately with AI.

In order for your content to rank—and actually help people—you need relevant, original content and data to share.

The combination of “scalable” keywords and great data is a force to be reckoned with.

With a little know-how, it’s possible to generate thousands (even millions) of pageviews from a single page template. Pretty cool.

Portrait of Louise Linehan

Post byLouise Linehan

Louise is a Content Marketer at Ahrefs. Over the past ten years, she has held senior content positions at SaaS brands: Pi Datametrics, BuzzSumo, and Cision. During this time, she has published hundreds of blog posts, spearheaded industry-leading research, and developed expert-led webinar programs.

Louise first learned the ropes of SEO during her time at enterprise SEO company, Pi. Within weeks of joining, she had published a piece on cannibalization that caught the attention of the biggest name in SEO at the time: Rand Fishkin. Since then, she has written on topics ranging from manipulative link building, to content syndication, and search demand forecasting. In her spare time, she has deepened her practical understanding of SEO by building websites for two local businesses.

You can read Louise's words in industry publications like Search Engine Watch, The Drum, The Telegraph, and Spin Sucks.

Keep learning


[Programmatic SEO, Explained for Beginners - 图31

Louise Linehan](https://ahrefs.com/blog/author/louise-linehan/)March 25, 2025

General SEO

### Google Is Looking Out for #1. It’s Time You Do, Too


[Programmatic SEO, Explained for Beginners - 图32

Si Quan Ong](https://ahrefs.com/blog/author/si-quan-ong/)May 13, 2024

General SEO

### 56 Google Search Statistics to Bookmark for 2024


[Programmatic SEO, Explained for Beginners - 图33

Michal Pecánek](https://ahrefs.com/blog/author/michal-pecanek/)January 14, 2020

Content Marketing

General SEO

[### Open Graph Meta Tags: Everything You Need to Know

Programmatic SEO, Explained for Beginners - 图34](https://ahrefs.com/blog/open-graph-meta-tags/)


[Programmatic SEO, Explained for Beginners - 图35

Mateusz Makosiewicz](https://ahrefs.com/blog/author/mateusz-makosiewicz/)November 11, 2021

General SEO

### SEO for Startups: How to Grow Search Visibility on a Budget


[Programmatic SEO, Explained for Beginners - 图36

Sam Underwood](https://ahrefs.com/blog/author/sam-underwood/)January 18, 2024

General SEO

### 14 Ways to Improve Ecommerce Product Pages for SEO

Your weekly marketing’s must-reads

Join 284K marketers for weekly news, useful reads, industry updates, and the memes you didn’t know you needed.

Explore what’s inside the newsletter →

Blog

The Beginner’s Guide to SEO ↗

A beginner’s guide to SEO basics, with resources to help you start ranking higher.


Product


© 2026 Ahrefs Pte. Ltd. (201227417H) 16 Raffles Quay, #33-03 Hong Leong Building, Singapore 048581

Legal info

Privacy Policy

Cookie Settings

Terms of Service

English