Fireflies.ai to Google Sheets, scored calls logged
Your calls are getting recorded. Great. Then the real mess starts: someone has to find the transcript, skim it, “score” it, and somehow turn that into coaching notes your team will actually use.
Sales managers feel it first because call review piles up fast. A customer success lead gets dragged into the same loop when onboarding calls slip. Even a founder ends up listening to random recordings at night. This Fireflies Sheets scoring automation turns every transcript into consistent, logged coaching data without the admin grind.
Below you’ll see exactly how the workflow runs in n8n, what outcomes you can expect, and what you need to set it up cleanly.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: Fireflies.ai to Google Sheets, scored calls logged
flowchart LR
subgraph sg0["Message a model Flow"]
direction LR
n0["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/webhook.dark.svg' width='40' height='40' /></div><br/>Webhook"]
n1@{ icon: "mdi:database", form: "rounded", label: "Append row in sheet", pos: "b", h: 48 }
n2@{ icon: "mdi:cog", form: "rounded", label: "Get a transcript", pos: "b", h: 48 }
n3@{ icon: "mdi:robot", form: "rounded", label: "Message a model", pos: "b", h: 48 }
n4["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/code.svg' width='40' height='40' /></div><br/>Code in JavaScript"]
n5["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/code.svg' width='40' height='40' /></div><br/>Code in JavaScript1"]
n6@{ icon: "mdi:message-outline", form: "rounded", label: "Send a message", pos: "b", h: 48 }
n7["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/slack.svg' width='40' height='40' /></div><br/>Send a message1"]
n0 --> n2
n3 --> n4
n2 --> n5
n4 --> n1
n1 --> n6
n1 --> n7
n5 --> n3
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 n3 ai
class n1 database
class n0 api
class n4,n5 code
classDef customIcon fill:none,stroke:none
class n0,n4,n5,n7 customIcon
The Problem: Call Reviews Don’t Scale (But Your Pipeline Does)
Manual call review is one of those “we’ll do it when we have time” habits that quietly wrecks performance. You might review a handful of calls when a rep is new, then the calendar fills up and suddenly weeks go by with zero feedback. Worse, scoring is inconsistent. One manager cares about discovery. Another cares about objection handling. So your team gets mixed messages, and your CRM notes become a dumping ground of half-finished thoughts. The cost isn’t only time. It’s lost deals, slower ramp, and coaching that happens too late to matter.
The friction compounds.
- Someone has to track down the right transcript link and paste it into whatever doc or spreadsheet you’re using.
- Call scoring changes depending on who listened, which means trends in “performance” are basically noise.
- Managers delay feedback because the review process takes too long to do “properly.”
- Insights live in Slack, email, and random notes, so reporting turns into a monthly scramble.
The Solution: Score Every Fireflies Transcript and Log It Automatically
This workflow uses n8n to turn each new Fireflies.ai transcript into a scored, searchable record in Google Sheets, with instant notifications for your team. It begins when a webhook receives the event that a transcript is ready. n8n then pulls the transcript details from Fireflies, cleans and formats the text so it’s ready for analysis, and sends it to OpenAI for an evaluation. Once the AI returns a structured score and notes, the workflow parses those results into neat fields. Finally, it writes a new row into your Google Sheet with lead and call details, then sends the evaluation to Slack and/or Gmail so the right people see it immediately. No hunting. No copy-paste. No “I’ll review this later.”
The workflow starts with a new transcript event, then moves through formatting and AI scoring, and ends with two outputs. You get a row in Google Sheets for reporting and coaching, plus a real-time alert in Slack or Gmail so action happens while the call is still fresh.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say your team runs 25 calls a week and you try to review even “just a quick sample” of them. If you spend about 10 minutes finding the transcript and another 15 minutes skimming and scoring, that’s roughly 10 hours a week gone. With this workflow, you spend maybe 2 minutes confirming the sheet columns once, then each call scores itself in the background and posts a Slack alert when it’s ready. You still listen to a few key calls, but you stop doing busywork on all of them.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- Fireflies.ai to provide transcripts and call metadata
- Google Sheets to store scores, links, and notes
- Slack or Gmail to send the evaluation to your team
- OpenAI API key (get it from your OpenAI dashboard)
Skill level: Intermediate. You’ll connect accounts, map a few fields, and test with one real transcript.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
A new transcript hits your webhook. Fireflies signals that a call transcript is ready, and n8n catches it immediately so you don’t have to monitor anything manually.
The transcript is retrieved and cleaned up. n8n pulls the full transcript data from Fireflies, then formats the text into something the AI can evaluate reliably (less noise, clearer structure).
OpenAI reviews the call and returns a score. The AI step evaluates the transcript against your criteria (quality, engagement, outcomes), then produces a structured response you can store and compare later.
Results are logged and shared. A new row is added to Google Sheets with the lead fields, transcript URL, and AI notes, and then a Slack message and/or Gmail email sends the summary to the people who need it.
You can easily modify the scoring rubric and the notification routing based on your needs. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Webhook Trigger
Set up the inbound webhook that starts the pipeline and passes the meeting ID to the workflow.
- Add and open Incoming Webhook Trigger.
- Set HTTP Method to
POST. - Set Path to
66273cf2-950f-42e5-a386-60d643868985. - Set Response Mode to
lastNodeand Response Data toallEntries. - Ensure your calling system posts
meetingIdin the request body, since downstream uses{{$json.body.meetingId}}.
{"meetingId":"YOUR_FIRELIES_ID"} before moving on.Step 2: Connect Fireflies and Retrieve the Transcript
Fetch the call transcript using Fireflies so it can be formatted and analyzed.
- Open Retrieve Transcript Data and set Transcript ID to
{{$json.body.meetingId}}. - Credential Required: Connect your firefliesApi credentials.
- Verify the node output includes
data.sentences,data.speakers, anddata.transcript_urlfor later mapping.
meetingId, Retrieve Transcript Data will return empty data and downstream nodes will fail.Step 3: Format the Transcript and Configure AI Review
Transform the transcript into structured text and submit it to the AI model for scoring.
- Open Format Transcript Text and keep the provided JavaScript Code as-is; it produces
transcript_formatted,transcript_plain, andmetafields. - Open AI Review Request and set the Model to
gpt-4o. - In AI Review Request messages, keep the prompt text and verify the transcript insertion uses the expression
{{ $json.Transcript }}. - Credential Required: Connect your openAiApi credentials.
- Open Parse AI Scores and keep the provided JavaScript Code to extract and normalize the JSON response.
transcript_formatted and transcript_plain, but AI Review Request references {{$json.Transcript}}. If your data doesn’t include Transcript, update the expression to the correct field (for example {{$json.transcript_formatted}}).Step 4: Configure Spreadsheet and Notifications
Write the AI scores to Google Sheets and alert your team via email and Slack.
- Open Add Row to Sheet and set Operation to
append. - Set Document to
1TcWkY4KVQfl7n5n9UyjiV590GdzgbCOrLvfb8d6FreAand Sheet toSheet1(gid0). - Confirm column mappings use the expressions provided, such as
{{$json.pain_points}},{{$json.overall_score_percent}}, and{{$('Retrieve Transcript Data').item.json.data.transcript_url}}. - Credential Required: Connect your googleSheetsOAuth2Api credentials.
- Open Email Notification Dispatch and set Send To to
[YOUR_EMAIL], Subject toGrading Call System, and Message toHey, your google sheet is ready! Go check it out!. - Credential Required: Connect your gmailOAuth2 credentials.
- Open Slack Alert Post and set Text to
Hey, your google sheet is ready! Go check it out!and Channel to[YOUR_ID]. - Credential Required: Connect your slackOAuth2Api credentials.
Step 5: Test and Activate Your Workflow
Validate the full pipeline from webhook to notifications before enabling it in production.
- Click Execute Workflow on Incoming Webhook Trigger and send a test POST with
{"meetingId":"YOUR_FIRELIES_ID"}. - Confirm Retrieve Transcript Data returns transcript content and Format Transcript Text outputs
transcript_formattedortranscript_plain. - Verify AI Review Request returns JSON and Parse AI Scores produces flattened score fields.
- Check your Google Sheet to see a new row appended by Add Row to Sheet, then verify both Email Notification Dispatch and Slack Alert Post are received.
- Toggle the workflow to Active once the end-to-end test is successful.
Common Gotchas
- Fireflies.ai credentials can expire or need specific permissions. If things break, check your Fireflies integration settings and transcript export permissions first.
- If you’re using Wait nodes or external rendering, processing times vary. Bump up the wait duration if downstream nodes fail on empty responses.
- Default prompts in AI nodes are generic. Add your brand voice early or you’ll be editing outputs forever.
Frequently Asked Questions
About 30–60 minutes if your accounts and sheet are ready.
No. You will mainly connect accounts and map fields in n8n. The included code steps are already built and usually don’t need changes.
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 OpenAI API usage, which is typically a few cents per scored call depending on transcript length.
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. Update the AI Review Request node prompt to match your rubric (for example, discovery quality, next steps clarity, objection handling), then adjust the Parse AI Scores code node to match the fields you want written into Google Sheets. Common customizations include separate scores by category, a “coach this call” flag when the score is low, and routing Slack alerts only to the rep’s manager.
Usually it’s expired credentials or missing transcript export permissions in Fireflies. Reconnect the Fireflies node in n8n, then confirm the transcript you’re testing actually exists and is accessible to that account. If it works sometimes and fails in batches, you may be hitting rate limits or pulling transcripts before Fireflies finishes processing them, so adding a short delay can help.
A lot. On n8n Cloud, it depends on your monthly execution limit, and on self-hosting it mainly depends on your server and how long each OpenAI request takes. In practice, most small teams run dozens or a few hundred scored calls a week without issues if the workflow is paced and your API limits are reasonable.
Often, yes, if you care about control and consistency. This workflow has multiple steps that benefit from branching, formatting, and parsing, and n8n handles that without turning each little action into a separate paid task. You also get the option to self-host, which matters once you’re scoring lots of calls. Zapier or Make can still be fine for a simple “new transcript → send to Slack” setup, but the moment you want structured scoring in a sheet, you’ll appreciate the flexibility here. Talk to an automation expert if you want a quick recommendation based on your call volume.
Once this is running, every call leaves a trail you can actually use. The workflow handles the repetitive part, so your coaching becomes timely, consistent, and easy to track.
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.
