Migrate Your UTM System with an AI Agent | UTM Mind
A step-by-step guide on how to use an AI agent to automatically migrate your existing UTM conventions from CSVs, text files, or raw links into a structured and governed system in UTM Mind.
Ditch the Spreadsheet: How to Migrate Your Entire UTM System with an AI Agent
Your team has years of UTM conventions, but they're scattered across spreadsheets, text files, and a thousand legacy links. Migrating this to a new, structured system feels like an impossible task. Until now. With UTM Mind's AI Assistant, you can feed your existing chaos directly to an AI agent and have it automatically build your new, governed UTM workspace.
This guide will show you how to use an AI agent to read your existing UTMs from CSVs, text files, or a list of URLs, and have it automatically populate your UTM Mind Attributes, Templates, and even generate validated links.
The Challenge: Migrating Legacy UTMs
Most marketing teams face the same problem. Naming is inconsistent — the same source might appear as google, Google, google_ads, or cpc depending on who created the link and when. The data is scattered across Excel sheets, Google Sheets, .txt files, or is only visible in the URLs themselves. And the thought of manually copying and pasting thousands of values into a new system is a non-starter.
This is where an AI agent with the right tools becomes a game-changer.
Your AI Migration Assistant
UTM Mind's AI Assistant (or any agent connected to our MCP Server) has the tools to ingest your unstructured data and turn it into a structured, governed system.
The key tools for migration are:
| Tool | What it does |
|---|---|
upsert_attribute | Adds new values to your Attribute lists, or updates existing ones |
create_template | Builds a new UTM Template |
create_rule_set | Creates validation rules to enforce your conventions |
create_links_batch | Generates multiple validated links at once |
Here's how to use them with your existing data.
Scenario 1: Migrating from a CSV or Spreadsheet
This is the most common scenario. You have a spreadsheet with columns for utm_source, utm_medium, utm_campaign, etc.
Step 1: Prepare Your Data
Export your spreadsheet as a .csv file. Make sure it has clear headers. A typical export might look like this:
"utm_source","utm_medium","utm_campaign","notes"
"google","cpc","q4_black_friday_sale","Main sales campaign"
"meta","paid_social","retargeting_video_views","Top of funnel"
"linkedin","sponsored_post","ebook_download_devops","Content marketing"
Step 2: Prompt Your AI Agent
Upload the CSV to your agent and give it a clear prompt:
"I've uploaded campaigns.csv. Using utm-mind, do the following: For each unique value in the utm_source, utm_medium, and utm_campaign columns, use the upsert_attribute tool to add them to the corresponding Attribute category in UTM Mind. Then create a new Template called 'Legacy Campaigns' and a Rule Set called 'Legacy Governance' that enforces utm_source and utm_medium must be one of the values from the CSV."
The agent will parse the CSV and make a series of upsert_attribute and create_rule_set calls, instantly migrating your conventions.
Scenario 2: Migrating from a Text File
Sometimes you just have a simple list of approved values in a .txt file — perhaps a document your team has maintained for years.
Example sources.txt:
google
meta
linkedin
twitter
bing
outbrain
Upload sources.txt to your agent and prompt it:
"Using utm-mind, for every line in the attached sources.txt file, use the upsert_attribute tool to add the value to the 'utm_source' attribute category."
In seconds, your utm_source dropdown in UTM Mind is populated with all your approved sources.
Scenario 3: Migrating from Existing Links
This is the most powerful scenario. You don't have a clean list — just a collection of historical URLs from your analytics platform, ad platform exports, or a spreadsheet of past campaigns.
Paste a list of up to 50–100 links into a text file or directly into the agent prompt, then instruct the agent:
"I've provided a list of 50 URLs. For each URL, extract the values for utm_source, utm_medium, and utm_campaign. Using utm-mind's upsert_attribute tool, add each unique extracted value to its corresponding attribute category. After processing all URLs, give me a summary of how many unique sources, mediums, and campaigns you added."
The agent will read each URL, parse the query parameters, and systematically build your Attribute library from your real-world data. What would have taken a human analyst a full day takes the agent a few minutes.
From Migration to Generation
Once your attributes are migrated, you can immediately start generating new, clean links. The agent already knows your conventions, so every new link it creates will be validated against the rules it just built.
"Great. Now, using utm-mind, generate a new link for the 'q4_black_friday_sale' campaign on Google CPC. The destination URL is https://acme.com/shop."
The agent uses the create_link tool, and because the values are already in your Attribute library, the link is instantly valid against your newly created rules.
The Complete Migration Workflow
Here's the end-to-end process in a single view:
| Step | Action | Tool Used |
|---|---|---|
| 1 | Export existing UTM data to CSV or TXT | — |
| 2 | Upload file to UTM Mind AI Assistant | — |
| 3 | Agent reads and extracts unique values | AI parsing |
| 4 | Agent populates Attribute categories | upsert_attribute |
| 5 | Agent creates a governing Template | create_template |
| 6 | Agent creates validation Rule Sets | create_rule_set |
| 7 | Start generating clean, validated links | create_link |
Start Your Automated Migration
Stop dreading the migration process. Connect your AI agent to UTM Mind, feed it your existing files, and let it do the heavy lifting. You can modernize years of messy UTMs in an afternoon.