Batch Keyword Generation: A Content Planning Strategy for Programmatic SEO

Last week, I spent three hours using Ahrefs to mine keywords, just to organize 200 long-tail terms. A friend told me he generated over 2,000 keyword matrices in a single day using a programmatic SEO approach. Where’s the gap?

It all comes down to the keyword batch generation workflow. Instead of writing content page by page, I systematically combine seed keywords, modifiers, and connectors to batch-produce content pages that match search intent.

This article will give you an actionable 5-step workflow: from seed keyword expansion to modifier mining, then to keyword matrix construction and intent classification. You can start implementing right after reading. If you’ve already read “What is Programmatic SEO: Application Boundaries and Anti-Spam Policy Red Lines”, this is the practical guide—specifically focusing on how to do batch keyword generation.

Core Logic of Batch Keyword Generation

The core logic of programmatic SEO keyword batch generation is actually quite simple—not the complex algorithm you might imagine.

It’s just one formula: [Seed Keyword] + [Modifier] + [Connector].

Seed keywords are your business’s core concepts, like “lawyer,” “insurance,” “loan.” Modifiers are subdivisions of the core term, like “divorce,” “criminal,” “civil”—these are mined from user needs. Connectors? Geographic locations (Beijing/Shanghai/Guangzhou), use cases (online/free/fast), price ranges (cheap/high-end).

Let me share a real case I’ve seen. A legal services content site chose “lawyer” as the seed keyword, then extracted “divorce lawyer,” “criminal lawyer,” and “property lawyer” as modifiers from user Q&A, and used over 300 city names nationwide as connectors. Combined, nearly 20,000 long-tail keywords. Each keyword corresponds to a landing page—“Beijing divorce lawyer fee standards,” “Shanghai criminal law firm recommendations”—all terms users actually search for.

This is the power of batch generation. Manual mining? You have to think of them one by one, fill them in one by one. Programmatic? Once your seed keywords, modifiers, and connectors are designed, the rest is permutation and combination. The efficiency gap isn’t 2x or 3x—it’s over 10x.

There’s an interesting data point: after implementing programmatic SEO strategies, blogs see an average traffic increase of 3.5x (source: bloghunter.se). This isn’t hype—it’s statistics from real sites. Of course, the prerequisite is doing it compliantly—don’t cross Google’s red lines. I’ll cover this risk point specifically later.

5-Step Keyword Batch Generation Workflow

Alright, logic covered. Now let me give you a workflow you can use right away—five steps total.

Step 1: Seed Keyword Mining

Seed keywords are the foundation of the entire system. Choose wrong, and subsequent modifiers and connectors are wasted.

How to find them? I use Ahrefs Keywords Explorer. Enter your industry’s core term, like “lawyer,” and look at search volume, relevance, and what terms competitors are using. SEMrush’s Keyword Magic Tool works too—similar functionality.

One principle: seed keywords should be “core” enough, but not too broad. “Legal services” is too broad—the search intent isn’t clear. “Lawyer” is good—when users search this term, they basically know what they’re looking for.

Step 2: Modifier Expansion

Where do modifiers come from? Three channels: UGC (user-generated content), competitor analysis, and search suggestions.

UGC is most reliable. Go to Zhihu, Xiaohongshu, industry forums, and see how users ask questions. “How much for a divorce lawyer in Beijing?” “Is it necessary to hire a lawyer for criminal cases?”—these questions hide modifiers. Divorce, criminal, property, traffic accidents—all mined from real user needs.

Competitor analysis works too. Open competitor websites and look at their navigation bars, category tags, landing page titles. What terms they use, you can reference—not copy, but learn the approach.

Search suggestions are a free shortcut. Type your seed keyword in Google’s search box and watch the dropdown suggestions. Under “lawyer,” you’ll see “law firm,” “lawyer consultation,” “lawyer fee standards”—all potential modifier candidates.

Step 3: Connector Combination

Connectors fall into three categories: geographic locations, use cases, and price ranges.

Geographic locations are easiest. Doing local services? Pull out all city names nationwide. I use Whitespark’s Local Keyword Tool, specifically for this—one-click export of 300+ city-level names, saving manual organization time.

Use cases? “Online,” “free,” “fast,” “professional”—when users search these terms, they have specific needs in mind. “Online lawyer consultation” and “lawyer consultation” have completely different search intents.

Price ranges matter too. “Cheap,” “high-end,” “good value”—these terms work particularly well in e-commerce and service industries. “Cheap divorce lawyer” might have low search volume, but low competition and high conversion rates.

Step 4: Keyword Matrix Construction

Now you have seed keywords, modifiers, and connectors. How do you combine them into a matrix?

I use Excel or Airtable. Create a table with columns: seed keyword, modifier, connector, combined keyword, search volume (estimated), difficulty (estimated).

Then use formulas to batch-generate combined keywords. Excel uses CONCATENATE; Airtable uses formula fields. In seconds, thousands of keywords appear.

Free tools exist too. Keyword Shitter (funny name) can batch-expand keywords—export and manually filter afterward. ChatGPT or Claude can help too—give them seed keywords and modifiers, let them combine for you, saving time.

Step 5: Search Intent Classification

Keywords are ready, but don’t use them all at once. You need to classify by search intent: informational, navigational, transactional.

Vultr - High-performance NVMe VPS with 32 global locations, one-click Docker deploy

View Pricing →Ad

Informational keywords: users are looking for answers. “Divorce lawyer fee standards”—they want to understand pricing, not ready to pay yet. These terms suit informational articles and Q&A pages.

Navigational keywords: users are looking for specific sites. “So-and-so law firm official website”—they have a target in mind, just unsure of the URL. These terms suit brand landing pages.

Transactional keywords: users are ready to act. “Beijing divorce lawyer appointment,” “online lawyer consultation payment”—they’re ready to spend. These terms are best for service landing pages.

How to judge? Look at the keyword itself. “How much,” “how about,” “what is”—mostly informational. “Appointment,” “buy,” “consult”—mostly transactional. SEOMatic’s article has detailed methodology—I’ve tried it, quite reliable.

Data Structuring Methodology

Keyword matrix is ready, next is data structuring. Many skip this step—thinking keywords are enough, why bother organizing data?

I’ve fallen into this trap. Before, I generated a bunch of keywords and jotted them down haphazardly in a notebook. Later, when I needed to batch-create pages, I found them unusable—some keywords had no search volume data, some were duplicates, some lacked intent classification. Going back to organize wasted time.

So data structuring is the prerequisite for scaling content.

Keyword Data Table Design

I use Airtable for tables—better than Excel, supporting multiple views, filtering, and linked fields.

Here’s the table structure:

Each field has its purpose. Search volume and difficulty determine priority—high volume, low difficulty terms go first. Intent classification determines page type—informational for articles, transactional for landing pages. Status field tracks progress—thousands of keywords without status management become chaos.

Data Normalization Principles

Four principles:

1. Unified Format: Store all keywords in unified format. “Beijing divorce lawyer” and “Beijing divorce lawyer” should be merged—don’t keep two versions.
2. Deduplication: Excel has deduplication; Airtable has duplicate detection. Batch generation easily produces duplicates—must clean them.
3. Classification: Classify by intent, priority, and status. When batch-creating pages later, just filter and use.
4. Source Attribution: Mark data source for each keyword—Ahrefs or SEMrush, UGC or competitor. Data sources have different credibility—need basis for judgment.

SEORAF has a free keyword research template—I’ve referenced it, structure design is quite reasonable. You can download and modify for your needs.

Content Template Design

Keyword data is ready, next is content template design.

Programmatic SEO’s core idea: you’re not writing content page by page, but batch-generating with templates. Templates have fixed structures and dynamic variables.

Core Elements of a Template

A qualified content template should include at least:

• Title: Formula-generated, like “{{City}}{{Service}}{{Keyword}}” → “Beijing Divorce Lawyer Fee Standards”
• Description: Meta Description template, like “Learn about {{City}}{{Service}} {{keyword}}, professional lawyers provide…”
• Content Structure: Fixed H1/H2/H3 hierarchy, variable content blocks
• Schema Markup: Structured data to help Google understand page content
• Internal Link Positions: Automatically insert related article links

Template Variable Design

Variables are the soul of templates. Use {{variable_name}} to mark dynamic content—automatically replaced when template is filled.

Example:

Title: {{City}} Divorce Lawyer Fee Standards - 2026 Latest Analysis

H1: {{City}} Divorce Lawyer Fee Standards

First paragraph: In {{City}}, divorce lawyer fees typically range between {{PriceRange}}...

After filling:

Title: Beijing Divorce Lawyer Fee Standards - 2026 Latest Analysis

H1: Beijing Divorce Lawyer Fee Standards

First paragraph: In Beijing, divorce lawyer fees typically range between 3000-10000 RMB...

Each page has the same structure but different content—this is the essence of programmatic SEO.

Railway - Modern deployment platform, zero-config, go live in minutes

Start Now →Ad

Differentiation Strategy

Template-generated content can be too similar. Google doesn’t like duplicate content. How to differentiate?

1. Data Differences: Different cities, different services have different prices, processes, and cases. Write these differences into templates.
2. Localization Details: Add local policies, local cases, local lawyer recommendations. Like “Beijing Civil Affairs Bureau divorce process,” “Shanghai court judgment cases.”
3. UGC Material: Extract real questions and answers from user Q&A, mix into template content.
4. Dynamic Updates: Regularly update data to keep pages fresh.

Real Case References

Two classic cases: Wise and Zapier.

Wise: Currency conversion pages. Their template is “{{Currency A}} to {{Currency B}} exchange rate.” Like “USD to CNY exchange rate,” “EUR to GBP exchange rate.” Each page has the same structure, but exchange rate data updates in real-time—content differentiation comes from data itself.

Zapier: Integration pages. Template is “{{Tool A}} and {{Tool B}} integration.” Like “Slack and Trello integration,” “Notion and Google Sheets integration.” Each page explains how two tools connect—content difference comes from the tools’ own characteristics.

Both sites use programmatic SEO, both have millions in traffic. The secret is good template design—unified structure, differentiated content, smooth user experience.

Mainstream Tool Comparison and Selection

Tool selection matters for keyword batch generation. Choose wrong—either insufficient functionality or wasted budget.

I’ve compared mainstream tools for your reference.

Tool Characteristics

Ahrefs: Strongest keyword mining capability, large database, accurate data. But weak batch generation—need to manually export data and combine. Good for seed keyword mining, not for batch generation.

SEMrush: Keyword Magic Tool is useful, supports batch export of combined keywords. Limited template design—can only do simple templates. Good for keyword expansion, not for complex programmatic SEO.

Whitespark: Specialized in local SEO, geographic modifier generation is top-notch. One-click export of city names, region names, landmark names—saves lots of time. If your business is local services, this tool is essential.

SEOMatic: Programmatic SEO specialized tool. Batch generation, template design, auto-publishing—all-in-one service. Weak keyword mining—need to import data from other tools. Good for scenarios where you already have keyword data and just want to quickly create pages.

Recommended Combinations

My suggestion: use combinations, don’t just buy one.

• Ahrefs + Whitespark + Airtable: Seed keywords with Ahrefs, geographic modifiers with Whitespark, data management with Airtable. Good for local service businesses.
• SEMrush + SEOMatic: Keyword expansion with SEMrush, page generation with SEOMatic. Good for quick scaling.
• Ahrefs + ChatGPT/Claude: Seed keywords with Ahrefs, modifier combination with AI assistance, manual data management. Low-cost solution for limited budgets.

How to allocate budget? Ahrefs is a must-buy—most comprehensive database. Whitespark is cheap, essential for local SEO. SEOMatic depends on needs—if you want high automation, can buy. Airtable is free, basic features are sufficient.

AI-Assisted Strategy for 2026

AI tools have developed rapidly in the past two years—keyword batch generation can also use AI assistance. But one principle: AI assists, doesn’t replace manual review.

What AI Can Do

Keyword Combination: Give ChatGPT or Claude your seed keywords and modifier lists, let them combine for you. Much faster than manually writing formulas.

Alibaba Cloud - Top choice for developers, exclusive 15% off for Lighthouse servers

Get 15% Discount →Ad

Intent Classification: Throw keyword lists to AI, let it judge each word’s search intent. “How much for a divorce lawyer”—informational or transactional? AI judgment is basically accurate, saving your time analyzing one by one.

Cluster Analysis: Thousands of keywords, how to classify? AI can help you cluster by topic—like grouping “divorce lawyer” related terms together, “criminal lawyer” related terms into another group.

Template Content Filling: Template structure is ready, use AI to fill blank blocks. Like “Beijing divorce lawyer fee standards” page, AI can write price ranges, process explanations, and notes for you.

What AI Can’t Do

Completely Generate Content: Don’t let AI one-click generate entire articles or entire pages. Google updated policy in 2024, cracking down hard on AI-generated low-quality content. You might cross the red line—I explained this risk in detail in “What is Programmatic SEO: Application Boundaries and Anti-Spam Policy Red Lines,” recommend reading it.

Replace Human Judgment: Data AI gives you must be manually verified. Search intent judgment might be wrong, price data might be outdated, case details might be fabricated. AI won’t tell you these issues itself.

Quality Control Process

My approach: AI assistance + manual review in three checkpoints.

1. First checkpoint: Data Verification. AI-given search volume and difficulty data—check again with Ahrefs or SEMrush to confirm accuracy.
2. Second checkpoint: Content Verification. AI-filled template content, read through manually. Is the logic right? Is data outdated? Any fabricated cases?
3. Third checkpoint: SEO Check. After pages are generated, crawl with Screaming Frog or Sitebulb to check if Title, Description, Schema are correct, any duplicate content.

Three checkpoints completed, then publish. Omnius’s article has detailed AI-assisted workflow—I’ve referenced it, method is reliable.

Summary

After all this, the core message is one thing: keyword batch generation is the first step of programmatic SEO—do it well and everything after becomes easier.

Let’s review the 5-step workflow: seed keyword mining → modifier expansion → connector combination → keyword matrix construction → search intent classification. Each step has specific tools and methods—not empty talk.

Data structuring is the prerequisite for scaling content—design tables well, define fields clearly, normalize data, then batch-creating pages will go smoothly.

Don’t agonize over tool selection: Ahrefs for seed keywords, Whitespark for geographic modifiers, Airtable for data management, SEOMatic for template generation. Use them in combination—best value.

AI assistance works, but don’t be lazy—three manual review checkpoints are mandatory, or you’ll regret crossing Google’s red lines.

Next step suggestion: follow this article’s Airtable template structure, try generating your first batch of keyword matrices for your site. Then read “What is Programmatic SEO: Application Boundaries and Anti-Spam Policy Red Lines” to understand compliance boundaries before executing.

References

• What is Programmatic SEO: Application Boundaries and Anti-Spam Policy Red Lines
• SEOMatic: Programmatic SEO Keyword Generation Guide
• Whitespark Local Keyword Tool
• SEORAF Keyword Research Template
• Omnius AI-Assisted Content Workflow
• Bloghunter.se programmatic SEO traffic growth statistics

This article is the second in the programmatic SEO series (methodology chapter). The previous article “What is Programmatic SEO: Application Boundaries and Anti-Spam Policy Red Lines” covers theoretical boundaries, this one covers practical workflow.

13 min read · Published on: Apr 1, 2026 · Modified on: Apr 1, 2026

Easton

Digital Media