+ Parameter Template (Filled Example) + Optional AEO and GEO Add-on Module

This is Part 1 of the Dataset-First SEO case series. Full case with step-by-step logic. If your team uses LLM to draft SEO content and the output swings from decent to unusable, the problem is the lack of locked requirements before drafting.

The fix: I have the LLM run SERP analysis on the top 10 pages, extract on-page element counts (headings, lists, tables, links, word count), and set explicit targets. These parameters become input for a prompt chain - the model follows the dataset, not its own assumptions. This template shows the 23 parameters I lock before drafting. Want to know how to get the LLM to fill this in? Ask me.

Copy-ready schema (23 inputs)

**CONTEXT**
{{WEBSITE_URL}}
{{TOPIC}}
{{WORD_COUNT}}
{{COMPANY_NAME}}
{{AUDIENCE_DESCRIPTION}}
{{LANGUAGE}}
{{CONTENT_TYPE}}

**META & SEO ELEMENTS**
{{H1}}
{{META_TITLE}}
{{META_DESCRIPTION}}
{{URL_SLUG}}

**KEYWORDS**
{{MAIN_KEYWORD}}
{{SUPPORTING_KEYWORDS}}
{{LSI_KEYWORDS}}
{{PRIMARY_LSI_KEYWORD}}

**STRUCTURE TARGETS**
Derived from SERP analysis of top 10 pages
{{NUM_H2}}
{{NUM_H3}}
{{NUM_H4}}
{{NUM_LISTS}}
{{NUM_TABLES}}
{{NUM_IMAGES}}
{{NUM_INTERNAL_LINKS}}.             
{{NUM_EXTERNAL_LINKS}}

Parameter Template (Filled Example)

Example values for a B2B SaaS support topic.

**CONTEXT** 
{{WEBSITE_URL}} <https://www.example-company.com/> {{TOPIC}}
AI Agent: Definition, Examples, and How It Works for Customer
Support
{{COMPANY_NAME}}
Example Company
{{AUDIENCE_DESCRIPTION}}
Customer support leaders and operations managers at mid-sized SaaS companies who want
to standardize support workflows and reduce ticket handling time with AI. {{LANGUAGE}}
English
{{PAGE_TYPE}} SEO-optimized long-
form blog article

**META & SEO ELEMENTS**
{{H1}} AI Agent: Definition, Examples, and How It Works for
Customer Support
{{META_TITLE}} AI Agent for Customer Support: Definition,
Examples, and Setup
{{META_DESCRIPTION}}
A practical guide to AI agents for support teams, including examples, workflows, and
a simple implementation checklist. {{URL_SLUG}} ai-agent-customer-support

**KEYWORDS**
{{MAIN_KEYWORD}}
AI agent
{{SUPPORTING_KEYWORDS}}
AI agent definition, AI agent example, how AI agents work, AI agent use cases, AI agent
for customer support, AI agent vs chatbot, implementing AI agents, human-in-the-loop
{{LSI_KEYWORDS}} ticket routing, escalation rules, knowledge base automation, workflow
orchestration, autonomous workflows, agentic systems, retrieval augmented generation,
support operations {{PRIMARY_LSI_KEYWORD}} ticket routing

**STRUCTURE TARGETS**
{{WORD_COUNT_TARGET}}
1800
{{NUM_H2_TARGET}}
8
{{NUM_H3_TARGET}}
5
{{NUM_H4_TARGET}}
0
{{NUM_LISTS_TARGET}}
5
{{NUM_TABLES_TARGET}}
1
{{NUM_IMAGES_TARGET}}
4
{{NUM_INTERNAL_LINKS_TARGET}}
4
{{NUM_EXTERNAL_LINKS_TARGET}}
5

Optional AEO and GEO Add-on Module

This parameter set is not part of the base case. At the time of the project, I was not optimizing specifically for AEO or GEO. However, the same dataset-first pipeline can be extended with an add-on module when the goal includes answer-first structure and citation-ready content.

AEO block (answer-first structure)
{{PRIMARY_QUESTION}}
{{SHORT_ANSWER}}
{{PAA_QUESTIONS}}
{{NUM_FAQ_QA}}
* PRIMARY_QUESTION: the main user question in query
form.
* SHORT_ANSWER: a 40 to 70 word answer placed right after the lead.
* PAA_QUESTIONS: a list of People Also Ask questions, or a curated substitute
ifcollected manually. * NUM_FAQ_QA: the number of Q and A items placed at the end or
throughout the page.
GEO block (citation-ready content for generative
answers)
{{ENTITIES}}
{{KEY_FACTS}}
{{CITED_SOURCES}}
{{NUM_EXTERNAL_CITATIONS}}
* ENTITIES: key entities that must appear explicitly (products, platforms,
standards,terms).
* KEY_FACTS: claims that should be phrased in a citable way (short,
checkablestatements).
* CITED_SOURCES: sources you want to reference, or at least source types if URLs
are notavailable. * NUM_EXTERNAL_CITATIONS: the count of external citations that
support specific claims.
Note: NUM_EXTERNAL_LINKS and NUM_EXTERNAL_CITATIONS are different. External links navigational. External citations support specific claims.
 navigational. External citations support specific claims.

Full case with step-by-step logic Part 1 Full case with step-by-step logic Part 2 The next step №2 Brief Template (blank): the schema I use to generate briefs, with placeholder slots +

№2.1 Meta-Prompt: It takes your filled parameters (№1) and your brief template (№2), and outputs a ready-to-run prompt with all placeholders replaced.