Telegram to Google Docs, build an ICP you can trust
Your targeting looks “fine” until it doesn’t. One person’s idea of the Ideal Customer Profile turns into three different definitions, and suddenly every campaign, list build, and outbound sequence starts drifting.
Growth marketers feel it when lead quality drops. Sales leaders see it as wasted follow-ups. And founders end up being the human tie-breaker. This Telegram ICP automation turns real customer LinkedIn profiles into a consistent ICP doc your team can actually align around.
You’ll see how the workflow takes LinkedIn URLs from a chat, enriches them, has AI synthesize patterns, then outputs a clean ICP, a scoring model, and a Boolean search string in Google Docs.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: Telegram to Google Docs, build an ICP you can trust
flowchart LR
subgraph sg0["When chat message received Flow"]
direction LR
n0@{ icon: "mdi:play-circle", form: "rounded", label: "When chat message received", pos: "b", h: 48 }
n1@{ icon: "mdi:robot", form: "rounded", label: "AI Agent", pos: "b", h: 48 }
n2@{ icon: "mdi:brain", form: "rounded", label: "Anthropic Chat Model", pos: "b", h: 48 }
n3@{ icon: "mdi:memory", form: "rounded", label: "Simple Memory", pos: "b", h: 48 }
n4@{ icon: "mdi:cog", form: "rounded", label: "Airtop Data Enrichment", pos: "b", h: 48 }
n5@{ icon: "mdi:cog", form: "rounded", label: "Google Docs", pos: "b", h: 48 }
n6@{ icon: "mdi:cog", form: "rounded", label: "Google Docs1", pos: "b", h: 48 }
n1 --> n5
n5 --> n6
n3 -.-> n1
n2 -.-> n1
n4 -.-> n1
n0 --> n1
end
%% Styling
classDef trigger fill:#e8f5e9,stroke:#388e3c,stroke-width:2px
classDef ai fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
classDef aiModel fill:#e8eaf6,stroke:#3f51b5,stroke-width:2px
classDef decision fill:#fff8e1,stroke:#f9a825,stroke-width:2px
classDef database fill:#fce4ec,stroke:#c2185b,stroke-width:2px
classDef api fill:#fff3e0,stroke:#e65100,stroke-width:2px
classDef code fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
classDef disabled stroke-dasharray: 5 5,opacity: 0.5
class n0 trigger
class n1 ai
class n2 aiModel
class n3 ai
class n5 disabled
class n6 disabled
The Problem: Your ICP Lives in People’s Heads
Most teams “have an ICP,” but it’s really a handful of half-true assumptions. One spreadsheet column says “job title,” a Slack thread says “mid-market,” and the last campaign brief says “B2B SaaS only.” Then someone pulls five “great customers” from memory, but they’re not representative, and the patterns you needed never surface. Worst part, you don’t notice the drift until weeks later when pipeline is full of polite no-shows. That’s time spent writing ads, building lists, and running calls for the wrong people.
It adds up fast. Here’s where it usually breaks down.
- You collect LinkedIn profiles in random places, so the same work gets repeated every month.
- Manual “pattern spotting” turns into opinions, which means targeting changes depending on who’s in the meeting.
- Without a scoring model, you can’t explain why a lead is high-fit, so SDRs either cherry-pick or spray and pray.
- Search strings for finding lookalikes are inconsistent, so list quality swings wildly from week to week.
The Solution: Telegram → Enriched Profiles → ICP Doc
This workflow gives you a repeatable way to build (and rebuild) an ICP using evidence, not vibes. You send one or more LinkedIn profile URLs of your best customers through Telegram. An AI agent reads those inputs, calls an enrichment tool (Airtop) to extract structured profile details like role history, skills, and company context, then keeps the session “memory” so the analysis stays consistent as you add more examples. From there, a language model synthesizes the common traits into a clear ICP definition, a practical scoring method for future prospects, and a Google Boolean string you can paste into search or LinkedIn. Finally, n8n creates a timestamped Google Doc and appends the finished analysis so it’s shared, saved, and easy to revisit.
The workflow starts with a Telegram chat message containing LinkedIn URLs. Airtop enriches each profile into clean fields the model can reason over. Then the AI agent produces the ICP output and pushes it into Google Docs automatically.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say you pick 12 “high-fit” customers and want to build a real ICP from their LinkedIn profiles. Manually, you’ll spend maybe 10 minutes per profile collecting notes (about 2 hours), then another hour trying to turn those notes into something your team can reuse. With this workflow, you paste the 12 URLs into Telegram in one go, wait for enrichment and analysis, and you get a Google Doc output in roughly 10–15 minutes of actual effort. The doc is ready to share the same day, not next week.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- Telegram to submit LinkedIn URLs quickly
- Google Docs to store and share the ICP output
- Airtop API credentials (get it from your Airtop dashboard)
Skill level: Intermediate. You will connect accounts, add credentials, and do a little light troubleshooting if LinkedIn access changes.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
Telegram message trigger. You send a chat message that includes one LinkedIn profile URL (or a batch of them). That single message kicks off the workflow.
Profile enrichment and normalization. The AI agent parses the URLs, then Airtop extracts structured fields from each profile so you’re not analyzing messy copy-paste text. Session memory keeps context as multiple profiles are processed.
ICP synthesis with an LLM. Using an Anthropic chat model, the workflow identifies shared traits across your best customers and turns them into an ICP definition, a scoring rubric, and a Boolean search string you can reuse.
Google Docs output. n8n creates a new Google Doc with a timestamped title, then appends the finished ICP content so it reads cleanly and stays searchable for the team.
You can easily modify the output format to match your internal playbooks, or change what gets enriched (for example, focusing more on company signals than individual skills) based on your needs. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Chat Trigger
This workflow starts when a chat message is received, which then feeds LinkedIn profile URLs into the analysis pipeline.
- Add and open Incoming Chat Trigger.
- Keep the default settings under Options (no additional configuration is required).
- Connect Incoming Chat Trigger to ICP Analysis Agent to match the execution flow.
Step 2: Connect Google Docs
Google Docs is used to create a new ICP document and then append the AI-generated content.
- Open Create ICP Document and enable the node if it is disabled.
- Credential Required: Connect your googleDocsOAuth2Api credentials.
- Set Title to
=ICP Profile {{ new Date().toISOString().slice(0,16).replace('T', '_').replace(/:/g, '-') }}. - Set Folder ID to
[YOUR_ID]. - Open Append ICP Content and enable the node if it is disabled.
- Credential Required: Connect your googleDocsOAuth2Api credentials.
- Set Operation to
updateand Document URL to={{ $json.id }}. - In Actions, add an Insert action with Text set to
={{ $('ICP Analysis Agent').item.json.output }}.
Step 3: Set Up the AI Analysis Agent
The agent orchestrates LinkedIn enrichment and produces the final ICP analysis using Anthropic and session memory.
- Open ICP Analysis Agent and review the System Message to ensure it matches your ICP analysis requirements.
- Open Anthropic Language Model and confirm the model is set to
claude-3-7-sonnet-20250219. - Credential Required: Connect your anthropicApi credentials on Anthropic Language Model.
- Ensure Anthropic Language Model is connected to ICP Analysis Agent via the ai_languageModel connection.
- Confirm Session Memory Buffer is connected to ICP Analysis Agent via the ai_memory connection (no credentials required).
- Open LinkedIn Enrichment Tool and verify the URL field uses
={{ /*n8n-auto-generated-fromAI-override*/ $fromAI('URL', ``, 'string') }}. - Credential Required: Connect your airtopApi credentials on LinkedIn Enrichment Tool.
- Make sure LinkedIn Enrichment Tool is connected to ICP Analysis Agent via the ai_tool connection.
Step 4: Configure Output Flow
The AI output is written to a newly created document in Google Docs.
- Connect ICP Analysis Agent to Create ICP Document.
- Connect Create ICP Document to Append ICP Content so the created document ID flows forward.
- Confirm the execution order: Incoming Chat Trigger → ICP Analysis Agent → Create ICP Document → Append ICP Content.
Step 5: Test and Activate Your Workflow
Run a manual test to verify LinkedIn URLs are enriched and the ICP report is written to Google Docs.
- Click Execute Workflow and send a chat message to Incoming Chat Trigger with one or more LinkedIn profile URLs.
- Confirm ICP Analysis Agent produces a structured ICP output in its response data.
- Verify a new document appears in your target Google Docs folder with the ICP content inserted.
- If everything works, toggle the workflow to Active for production use.
Common Gotchas
- Airtop credentials can expire or need specific permissions. If things break, check your Airtop API key status and LinkedIn-connected profile inside Airtop first.
- If you’re enriching lots of profiles at once, processing times vary. Give the workflow more time between batches if downstream steps fail because a profile response came back empty.
- AI prompts that don’t include your brand context will produce generic ICP language. Add your positioning, exclusions, and “we never sell to…” notes early, honestly, or you will keep editing every run.
Frequently Asked Questions
About 30 minutes once your accounts are ready.
No. You’ll connect Telegram, Airtop, and Google Docs, then paste URLs to test it.
Yes. n8n has a free self-hosted option and a free trial on n8n Cloud. Cloud plans start at $20/month for higher volume. You’ll also need to factor in Airtop and model usage costs.
Two options: n8n Cloud (managed, easiest setup) or self-hosting on a VPS. For self-hosting, Hostinger VPS is affordable and handles n8n well. Self-hosting gives you unlimited executions but requires basic server management.
Yes, and you should. Most teams duplicate the workflow per segment, then adjust the ICP instructions inside the ICP Analysis Agent and tighten what the model should treat as “high-fit.” You can also change what Airtop extracts (for example, emphasize company size and industry signals) and tweak the Google Docs formatting in the append step. If you want separate scoring rubrics, keep the doc output the same but change the scoring section wording so it’s measurable.
Usually it’s a revoked or outdated Telegram credential in n8n, so reconnect the account and re-test the chat trigger. If Telegram is fine, the message format can be the issue too: send plain LinkedIn profile URLs (not shortened links) so the agent can parse them reliably. Also check that your n8n instance can receive webhook-style updates if you’re behind a firewall.
Dozens per run is realistic, and you can batch more if you’re patient with enrichment time.
Often, yes, because this workflow benefits from session memory, branching, and more control over how enrichment and AI steps run. Zapier and Make can handle “URL in, doc out,” but they get awkward when you want multi-profile batching, stateful analysis, and custom logic without hitting premium feature walls. n8n is also easier to self-host, which matters when you’re processing lots of profiles and don’t want per-task pricing surprises. If you only run this once a quarter, simpler tools may be fine. If it’s part of your weekly growth rhythm, n8n tends to fit better. Talk to an automation expert if you’re not sure which fits.
Once this is in place, your ICP stops being a debate and becomes an asset your whole team can use. Set it up, run it when you onboard new “great customers,” and keep targeting tight.
Need Help Setting This Up?
Our automation experts can build and customize this workflow for your specific needs. Free 15-minute consultation—no commitment required.
