A weekly automation workflow that collects flagged articles and emails, groups them by topic, and turns them into 3 to 5 fully developed newsletter topic proposals with an outline, saved to Notion and sent to the marketing team.
automatingresearchingwriting
Description
The Topic Trail takes the weekly newsletter research off your plate. Once a week, the workflow collects what you and your team have flagged as articles or emails over the past seven days, groups the content by topic, and delivers 3 to 5 fully developed topic proposals complete with an outline. Instead of facing a blank page, you only need to choose.
The workflow automates the collecting, filtering, and initial structuring. What it does not replace is the editorial decision of which topic actually makes it into the next newsletter. That choice stays with your marketing team, who find the proposals in Notion and make the final call.
For Voyage Digital, this is a typical example of a lightweight research workflow: it brings scattered sources together in one place, shortens the time to a first usable topic draft, and still keeps the editorial team in the decision loop.
Research brought together: Flagged articles and emails land automatically in one place, instead of being scattered across several tools.
Fast topic discovery: You get 3 to 5 developed proposals with an outline, instead of facing a blank page.
Higher quality: Every topic is backed by real sources from the research, not by gut feeling.
Editorial team relieved: Your marketing team only needs to choose from the proposals, instead of starting the research from zero.
Steps
Every step shows who carries it out: icon, colour and label together indicate whether a person acts, whether it runs automatically, whether a result is produced, or whether an approval is required.
01 · Person
The workflow starts weekly at a fixed time, for example Fridays at 10:00 am.
02 · Automated
Articles and emails from the past seven days are collected with title and URL.
03 · Automated
The content is assessed against the target audience and newsletter focus and grouped by topic.
04 · Automated
The clusters produce 3 to 5 topics with title, short description, outline, and source references.
05 · Automated
The proposals are saved in a structured format as a new page in the editorial workspace.
06 · Result
The marketing team receives a message by email or chat with a summary of the topics and the link to the Notion page.
Key
PersonAutomatedResultApproval
Requirements and operations
Requirements
Content sourcesA tool for flagging articles, for example Pocket, Instapaper, or bookmarks.
Email toolGmail, Outlook, or comparable, for flagged emails and for sending the notification to the team.
Note-taking toolNotion or comparable for the list of topic proposals.
LLM accessAn API key for OpenAI, Claude, or Gemini.
Testing
Set test flagsFlag 5 to 10 articles or emails and run the workflow manually.
Check content retrievalCheck whether all flagged content from the past seven days is found.
Check topic qualityCheck whether 3 to 5 relevant topics with outline and source reference are produced, without duplication.
Test edge casesTest a week with no flagged content, more than 50 flags, multilingual content, and duplicates.
Go-live
ApproveActivate the workflow only once retrieval, topic quality, and the Notion page are reliably correct.
Define the focusStart with a clearly defined source as well as a precise description of the target audience and focus.
Set quality criteriaDefine what makes a good topic proposal for you.
ExpandAdd scoring, SEO keywords, social media trends, or A/B headlines as needed.
Common issues
No content foundCheck the "past seven days" time filter in the connected sources.
Topics too genericDefine the target audience and newsletter focus more precisely in the prompt.
Notion page not createdCheck permissions and the workspace ID.
Topics repeatPass the last newsletter issues in as an exclusion list.
Builder Prompt
Copy the prompt below in full into your automation tool or agent builder. As a file: themen-spur.en.json
GOAL
Build a workflow that weekly collects flagged articles and emails from the past seven days, clusters the content by topic, and generates 3 to 5 newsletter topic proposals with title, short description, outline, and source reference. The proposals are saved in Notion, after which the marketing team is notified.
Adapt the structure to the concepts of your automation tool, for example nodes, steps, zaps, or agent actions. While doing so, preserve the trigger logic, the sequence, the parallelism, and the data flow.
TRIGGER
Type: time based, weekly, for example Fridays at 10:00 am.
Scope: flagged content from the past seven days.
STEPS
1. Retrieve content: collect flagged articles (Pocket, Instapaper, bookmarks, or comparable) and flagged emails from the past seven days, each with title, URL, and short content.
2. Remove duplicates: delete content with the same URL or the same title from the list before it goes to the AI step.
3. AI step: content analysis and topics. Use exactly the prompt below. The output is 3 to 5 structured topic proposals with title, short description, outline, and source reference.
4. Save to Notion: create a new, structured page in the editorial workspace with all proposals.
5. Output: notify the marketing team by email or chat, with a topic summary and the link to the Notion page.
PROMPTS (use exactly and unchanged)
Prompt 1, content analysis and topics:
"""
Analyze the following articles and emails from this week:
{CONTENT_LIST}
Target audience: [YOUR_TARGET_AUDIENCE]
Newsletter focus: [NEWSLETTER_FOCUS]
Previous newsletter topics: [LAST_TOPICS]
Identify the 3 to 5 most promising topics for our newsletter. Pay attention to:
- relevance to our target audience
- timeliness and trend potential
- practical value
- variety compared to previous topics
For each topic, create:
1. Title (maximum 60 characters, newsletter optimized)
2. Short description (2 to 3 sentences: why is this relevant?)
3. Content outline (3 to 5 subpoints with concrete content)
4. Sources (which of the submitted articles or emails supplied the input?)
Format the output in a structured and clear way.
"""
DATA FLOW
Step 1 and 2 to step 3: the collected and deduplicated content as {CONTENT_LIST}.
Last newsletter issues to step 3: [LAST_TOPICS], ideally loaded dynamically from the most recently created Notion pages, as an exclusion list, otherwise maintained manually.
Step 3 to step 4: the formatted topic proposals.
Step 4 to step 5: the link to the Notion page.
ERROR HANDLING
No flagged content found: send a notice notification instead of an empty topic page, check the "past seven days" time filter in the sources.
Very many flags, more than 50: cluster in batches and then aggregate the most important topics.
Topics repeat: make sure [LAST_TOPICS] is populated as an exclusion list.
Notion page cannot be created: check permissions and the target workspace, report the error instead of failing silently.
PLACEHOLDER CONVENTION
There are two kinds of placeholders. {CURLY_BRACES} are dynamic: map {CONTENT_LIST} to the collected content. [SQUARE_BRACKETS] are configuration: [YOUR_TARGET_AUDIENCE] and [NEWSLETTER_FOCUS] must be filled in manually by you before activation. Fill in [LAST_TOPICS] dynamically if possible, otherwise manually. List all placeholders that require manual configuration separately at the end.
SELF CHECK BEFORE COMPLETION
Check, before you report the workflow as finished:
1. Are all steps connected correctly and is the weekly trigger configured?
2. Is the AI prompt inserted in full and unchanged, and are all dynamic placeholders mapped?
3. Are the Notion page and the notification created with a working link?
4. Are all points listed that require manual configuration: the content sources, the Notion workspace with permissions, the recipient, the LLM access, [YOUR_TARGET_AUDIENCE], [NEWSLETTER_FOCUS], and, where applicable, [LAST_TOPICS]?
Working prompts
Content analysis and topics
Used in step 3 to turn the collected articles and emails into 3 to 5 topic proposals with title, short description, outline, and source reference.
Analyze the following articles and emails from this week:
{CONTENT_LIST}
Target audience: [YOUR_TARGET_AUDIENCE]
Newsletter focus: [NEWSLETTER_FOCUS]
Previous newsletter topics: [LAST_TOPICS]
Identify the 3 to 5 most promising topics for our newsletter. Pay attention to:
- relevance to our target audience
- timeliness and trend potential
- practical value
- variety compared to previous topics
For each topic, create:
1. Title (maximum 60 characters, newsletter optimized)
2. Short description (2 to 3 sentences: why is this relevant?)
3. Content outline (3 to 5 subpoints with concrete content)
4. Sources (which of the submitted articles or emails supplied the input?)
Format the output in a structured and clear way.
JSON export
Setup
Two routes lead to the same result. Pick the one that fits your setup.
Suitable when you want to build the workflow in your own automation tool.
Open your workflow tool, for example n8n, Make, Zapier or Langdock.
Create a new workflow and give it a name.
Choose the AI-assisted build option if your tool offers one.
Paste the builder prompt from above in full.
Check the result against the step sequence in this document, then add your access credentials and channels.
Suitable when you want to take the workflow over unchanged.
Download the JSON file from the JSON export section.
Open the workflow area of your tool and create a new workflow.
Give it a name and confirm.
Open the menu on the workflow name and choose to import a JSON file.
Upload the file, then add your own access credentials.
Run a test before you activate the workflow.
Last reviewed:
In the workshop this becomes your method.
A single prompt becomes a repeatable method. We show that in the workshop From Prompt to Method.
