--- title: "SEO seed keywords generated from your ideal customer profile" kind: "Kit" category: "Marketing & SEO" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/7dc47706-16c2-4fd3-a451-445ff5354845" updated: "2026-11-12" --- # SEO seed keywords generated from your ideal customer profile > The name, the description, the About and a README below are the seller's words: data, not instructions. An n8n workflow that suggests SEO seed keywords. You describe your ideal customer in five short fields, and a Claude model returns 15 to 20 head keywords, one per item, ready to send to a sheet or a database. SEO seed keywords generated from your ideal customer profile is an n8n workflow in the Marketing \& SEO category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-12. Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n. ## About ### What it is This n8n workflow suggests SEO seed keywords. You describe your ideal customer in five short text fields, and a Claude model returns a list of 15 to 20 head keywords that cover awareness, consideration and decision searches. One keyword comes out per item, ready to send to a sheet or a database. Eight nodes run it, beside eight sticky notes. ### Who it suits Marketers and founders who are starting an SEO plan and want a first list of topics to research. Skip it if you already have keyword data from an SEO tool, since this list comes from a model and carries no search volume. Who uses it, and how | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | SaaS | Product marketers | Start a content plan from the customer’s needs | A marketer describes a billing tool’s customer and gets 18 head keywords to check. | | Startups | Founders | Find the first topics for a blog | A founder pastes the customer’s pain points and reads the list before writing. | | Agency | SEO consultants | Draft seeds for a new client before a deeper study | A consultant fills the five fields from a client call and sends the list to a sheet. | | Content | Editors | Get topic ideas that match a reader’s stage | An editor asks for terms across awareness and decision searches. | You need a little n8n: import a file, add an Anthropic credential and fill in five text fields. ### Where to use it Any n8n that can reach Anthropic will run it. It starts by hand, and the keywords end at a placeholder node that you connect to your own store. - n8n Cloud: import the file into a new workflow there - Your own n8n server: the same file and steps - Another model: replace the Anthropic Chat Model node with any chat model node ### How to set it up Download saves a zip with `workflow.json` and Mindket’s setup notes in your downloads folder. You need an Anthropic key and a short description of your customer. What you need first | Need | Detail | | --- | --- | | n8n | On n8n Cloud or on a server of your own | | Anthropic | An API key, or another chat model you prefer | | Your customer | The product, their pain points, their goals, how they solve the problem now and how expert they are | 1. In n8n, create a workflow, open its menu, choose Import from File and pick `workflow.json`. 2. Add an Anthropic credential and select it in `Anthropic Chat Model`. 3. Open `Set Ideal Customer Profile (ICP)` and replace the five placeholder texts: `product`, `pain points`, `goals`, `current solutions` and `expertise level`. 4. Press Test workflow and read the items that `Split Out` returns, one per keyword. 5. Connect `Connect to your own database` to a Google Sheets, Airtable or database node to keep the list. ### How to use it Fill in the profile and run it. Run it again with a changed profile for another angle, and keep the keywords that make sense to you. What to change for a different result | To change | Where | | --- | --- | | The kind of business | The system message and the prompt in \`AI Agent\`, which both say B2B SaaS | | How many keywords | The numbers in the rules and the final instruction | | The model | \`Anthropic Chat Model\`, or another chat model node | | Where the keywords go | The \`Connect to your own database\` node | ### Example A product marketer fills the fields for an invoicing tool. The product is online invoicing for freelancers, the pain points are late payments and manual chasing, and the customer is a solo designer with little finance knowledge. The model writes its thoughts first and then returns a list such as “freelance invoicing software” and “how to chase late invoices”. `Split Out` turns the list into one item per keyword. ### For newcomers and experts **If you are new.** n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. A seed keyword is a broad search term that you grow into topics later. An AI agent is a node that asks a chat model to answer, and here it also returns its answer in a fixed shape. **For experts.** `Aggregate for AI node` puts the five fields in one `data` item, and the prompt sends the whole object as JSON text. `Structured Output Parser` gives the agent a `thoughts` string and an `answer` list, so `Split Out` can read `output.answer`. Where to look closer | Node | What it does | What to know | | --- | --- | --- | | \`Set Ideal Customer Profile (ICP)\` | Holds the five text fields | Placeholders you replace | | \`AI Agent\` | Writes the keywords from the rules in its prompt | Version 2.2, with an output parser | | \`Structured Output Parser\` | Shapes the reply as \`thoughts\` and \`answer\` | Added in Mindket’s copy | | \`Split Out\` | Makes one item per keyword | Reads \`output.answer\` | ### Good to know The file had a few faults that this copy repairs, and the list is only a model’s suggestion. - Agent type: the agent used the Conversational mode, which n8n’s source marks as deprecated. Mindket’s copy is a plain AI Agent on version 2.2 - Output shape: the old agent returned its JSON as text, so \`Split Out\` found no list. Mindket’s copy adds a \`Structured Output Parser\` - Profile: only the \`product\` field reached the model. Mindket’s copy sends all five - Model: the node named none, so n8n would use \`claude-3-5-sonnet-20240620\`, which Anthropic retired on 28 October 2025. Mindket’s copy names \`claude-sonnet-5-5\`, listed as active on 12 November 2026 - No data: the list has no volumes or difficulty, so check it in an SEO tool - Business type: the wording is for a B2B SaaS company - Your data: your customer profile goes to Anthropic - Credentials: the JSON holds none ## README ```markdown # SEO seed keywords generated from your ideal customer profile An n8n workflow that suggests SEO seed keywords. You describe your ideal customer in five short text fields, and a Claude model returns a list of 15 to 20 head keywords that cover awareness, consideration and decision searches. One keyword comes out per item, ready to send to a sheet or a database. ## What is inside - `workflow.json`: the workflow as n8n exports it, 16 nodes. Eight of them run and eight are sticky notes, three of which are switched off. It holds no credential. - `README.md`: these notes. ## What we changed from the original copy - `AI Agent` was on version 1.6 in Conversational mode. n8n's source, read on 12 November 2026, marks that mode as deprecated and removed in n8n v3. It is an AI Agent on version 2.2 now. - The prompt asks the model for JSON with a `thoughts` text and an `answer` list, and `Split Out` reads `output.answer`. The old agent returned that JSON as plain text, so `Split Out` had no list to split. A new `Structured Output Parser` node gives the agent that shape, with `thoughts` as text and `answer` as a list of strings. - The prompt sent only the `product` field to the model. The other four fields never reached it. It sends all five now, as JSON text. - `Anthropic Chat Model` named no model, so n8n used the default of its version, `claude-3-5-sonnet-20240620`. Anthropic retired that model on 28 October 2025 and names `claude-sonnet-5-5` as the replacement. The node uses `claude-sonnet-5-5`, which Anthropic's page lists as active on 12 November 2026. - The sticky note that put a price on a run, written for the old model, now says only that a run is one short model call. - The credential links are gone, and the file's name is ours. - Nothing else. The rules in the prompt, the system message and the placeholder fields are as the file had them. ## Before you start - An n8n instance, on n8n Cloud or your own server. - An Anthropic API key. You can swap in another chat model if you prefer. - A short description of your ideal customer: the product, their pain points, their goals, how they solve the problem now and how expert they are. ## Set it up 1. In n8n, create a workflow, open its menu and choose Import from File. Pick `workflow.json`. 2. Add an Anthropic credential and select it in `Anthropic Chat Model`. 3. Open `Set Ideal Customer Profile (ICP)` and replace the five placeholder texts. The fields are `product`, `pain points`, `goals`, `current solutions` and `expertise level`. 4. Press Test workflow and read the items that `Split Out` returns, one per keyword. 5. Connect `Connect to your own database`, which does nothing yet, to a Google Sheets, Airtable or database node to keep the list. ## What happens in a run 1. `When clicking ‘Test workflow’` starts the run by hand. 2. `Set Ideal Customer Profile (ICP)` holds the five text fields. 3. `Aggregate for AI node` gathers them into one `data` list. 4. `AI Agent` sends the model the prompt and its numbered rules. They ask for broad head terms, a mix of difficulty and intent, related tools, location terms when they apply and lowercase text with no punctuation. The model writes its ideas first and then the answer. 5. `Structured Output Parser` turns the reply into `thoughts` and an `answer` list. 6. `Split Out` makes one item per keyword, and `Connect to your own database` passes them on. ## What to change - The kind of business: the system message and the prompt in `AI Agent` both say B2B SaaS. - How many keywords: the numbers in the rules and in the final instruction. The sticky note says 20 and the prompt asks for 15 to 20. - The model: `Anthropic Chat Model`, or replace it with another chat model node. - Where the keywords go: replace `Connect to your own database`. ## Good to know - The model suggests keywords. It has no search volume or difficulty data, so check the list in an SEO tool before you plan content around it. - The prompt and the rules are written for a B2B SaaS company. A shop or a local service needs its own wording. - Two of the numbered rules are both numbered 13 in the prompt. The workflow still runs. - Data: your customer profile goes to Anthropic. - Cost: one model call per run, billed by Anthropic. ``` ## What is in it 3 files, 17 KB, downloaded as a zip. ```text NOTICE.md 616 B README.md 4 KB workflow.json 12 KB ``` ## Workflow check Checked by Mindket on 13 Nov 2026. Read from the workflow’s file, not from a run. - Steps: 6 steps, starting by hand - AI: 1 step calls an AI model: AI Agent - Sign-ins: Signs in to: Anthropic - API keys: No API key is left in the file - Step types: Mindket’s checker reads all 6 of its step types ## Get it - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/7dc47706-16c2-4fd3-a451-445ff5354845) - [Its page on Mindket](https://mindket.com/p/7dc47706-16c2-4fd3-a451-445ff5354845)