Trustpilot to Salesforce, sentiment notes that stay fresh
You get a new Trustpilot review, and then the busywork starts. Someone screenshots it, someone else pastes it into a CRM note, and a week later nobody can remember what customers were actually upset about.
This Trustpilot Salesforce automation hits Revenue Ops and Customer Success first, honestly, but marketing leads feel it too when positioning doesn’t match real feedback. You end up with stale account notes, missed patterns, and a lot of “we should really track this” conversations that go nowhere.
This workflow pulls review text from Trustpilot, summarizes it with AI, and updates both Salesforce and Google Sheets so the insight is always current. You’ll see what it does, what you need, and how teams use it to spot trends faster.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: Trustpilot to Salesforce, sentiment notes that stay fresh
flowchart LR
subgraph sg0["When clicking ‘Execute workflow’ Flow"]
direction LR
n0@{ icon: "mdi:play-circle", form: "rounded", label: "When clicking ‘Execute workf..", pos: "b", h: 48 }
n1["<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"]
n2@{ icon: "mdi:robot", form: "rounded", label: "AI Agent", pos: "b", h: 48 }
n3["<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"]
n4@{ icon: "mdi:database", form: "rounded", label: "Get row(s) in sheet", pos: "b", h: 48 }
n5@{ icon: "mdi:database", form: "rounded", label: "Append row in sheet", pos: "b", h: 48 }
n6@{ icon: "mdi:cog", form: "rounded", label: "Decodo", 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/salesforce.svg' width='40' height='40' /></div><br/>Create or update an account"]
n8@{ icon: "mdi:brain", form: "rounded", label: "OpenAI Chat Model", pos: "b", h: 48 }
n9["<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 JavaScript2"]
n10@{ icon: "mdi:swap-horizontal", form: "rounded", label: "If", pos: "b", h: 48 }
n11@{ icon: "mdi:swap-vertical", form: "rounded", label: "Loop Over Items", pos: "b", h: 48 }
n10 --> n9
n10 --> n11
n6 --> n1
n2 --> n3
n11 --> n6
n8 -.-> n2
n1 --> n2
n5 --> n10
n3 --> n5
n9 --> n7
n4 --> n11
n7 --> n11
n0 --> n4
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 n2 ai
class n8 aiModel
class n10 decision
class n4,n5 database
class n1,n3,n9 code
classDef customIcon fill:none,stroke:none
class n1,n3,n7,n9 customIcon
The Problem: Trustpilot Feedback Lives Outside Your CRM
Trustpilot is where customers say the quiet part out loud. The problem is that the feedback rarely shows up where your team actually works, which is usually Salesforce, a spreadsheet, or both. So reviews get checked “when someone has time,” copied into notes with zero structure, and then forgotten. Meanwhile, account owners are walking into renewals without the latest sentiment, and marketers keep guessing at themes that are already obvious in the reviews. The worst part is the drift: what customers complained about last month is not what they complain about this week.
The friction compounds. Here’s where it breaks down.
- Reviews are read manually, which means you only see them when you remember to look.
- CRM notes get written once and go stale, so reps rely on memory instead of data.
- Different people summarize reviews differently, and you lose consistent keywords and trends.
- When leadership asks “what’s driving sentiment lately?”, you’re stuck searching and scrolling.
The Solution: Trustpilot Reviews Automatically Enriched and Synced
This workflow turns Trustpilot reviews into up-to-date account intelligence inside Salesforce, plus a clean dataset in Google Sheets for reporting. It starts with a simple input list: Trustpilot review URLs paired with Salesforce Account IDs in a Google Sheet. n8n grabs each row in batches, scrapes the review content reliably using Decodo, and cleans the HTML so the AI model receives readable text. Then an AI Agent powered by an OpenAI chat model summarizes sentiment, pulls out themes, and extracts the “what’s going well / what’s going wrong” signals people actually need. Finally, the workflow writes the structured output into Google Sheets and, if a valid Account ID exists, updates a custom text field on the matching Salesforce Account with a trimmed, CRM-safe trend summary.
The workflow begins when you run it in n8n. It loops through your sheet rows, scrapes each Trustpilot page, and generates insights with AI. At the end, you get two synchronized outputs: a spreadsheet you can chart, and Salesforce notes that stay current.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say you track 30 Trustpilot URLs across key accounts, and you want updates each week. Manually, it’s maybe 5 minutes to open, skim, and paste notes per review, plus another 2 minutes to tidy wording for Salesforce, which is about 3 to 4 hours of work. With this workflow, you paste URLs and Account IDs into Google Sheets once, run it, and let n8n process the batch in the background while it writes the trend summary and the analytics row for you. Most teams get those same insights with only a quick spot-check at the end.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- Google Sheets to store URLs, IDs, and outputs.
- Salesforce to write sentiment into an Account field.
- Decodo API key (get it from your Decodo dashboard).
- OpenAI API key (get it from the OpenAI API settings page).
Skill level: Intermediate. You’ll connect accounts, add API keys, and create a custom Salesforce field.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
You run the workflow, and n8n pulls your input list. It starts from a manual launch, then reads rows from Google Sheets containing Trustpilot URLs and Salesforce Account IDs.
Reviews are scraped and cleaned. Each row is processed in batches, Decodo fetches the Trustpilot page content, and a cleaning step strips HTML so you’re left with usable review text.
AI generates consistent insight. An AI Agent with an OpenAI chat model summarizes sentiment, surfaces trends, and produces structured fields (themes, positives, negatives) so the output stays comparable week to week.
Two outputs are written in parallel. The workflow appends a row to Google Sheets for analytics, checks if a CRM ID exists, trims the trend summary to fit a Salesforce text field, and then updates the existing Account.
You can easily modify the summary format to match your team’s language based on your needs. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Manual Trigger
Start the workflow with a manual trigger so you can run and verify the review extraction process on demand.
- Add the Manual Launch Trigger node as the workflow entry point.
- Connect Manual Launch Trigger to Fetch Sheet Rows to kick off data ingestion.
Step 2: Connect Google Sheets
Pull review page URLs from your input sheet, then later append the AI-derived KPIs to an output sheet.
- Open Fetch Sheet Rows and set Document to
[YOUR_ID]. - Set Sheet Name to
gid=0(cached name:input). - Credential Required: Connect your googleSheetsOAuth2Api credentials.
- Open Append Sheet Record and set Document to
[YOUR_ID]. - Set Sheet Name to
697979134(cached name:output). - Confirm the Columns mapping uses the expressions shown in the node, such as
{{ $json.source_url }}and{{ $json.average_rating }}. - Credential Required: Connect your googleSheetsOAuth2Api credentials.
Tip: Ensure your input sheet includes a url column and (optionally) a salesforce_account_id column because several expressions reference {{ $('Fetch Sheet Rows').item.json.url }} and {{ $('Fetch Sheet Rows').item.json.salesforce_account_id }}.
Step 3: Set Up Scraping and Cleaning
Iterate through each row, scrape the review page, and convert HTML into clean text for analysis.
- Configure Batch Items Iterator to process rows in batches (default settings are fine).
- In Decodo Scrape Task, set URL to
{{ $('Fetch Sheet Rows').item.json.url }}. - Credential Required: Connect your decodoApi credentials.
- Confirm Decodo Scrape Task outputs to HTML Text Cleaner to generate cleaned text, entity slug, and metadata.
Step 4: Set Up AI Review KPI Extraction
Use the AI agent to analyze the cleaned review text and return structured KPIs.
- Open Review Insight Agent and keep the prompt as defined to enforce the JSON schema.
- Verify the prompt uses expressions like
{{ $json["url"] || "" }},{{ $json["domain"] || "" }}, and{{ $json["page_text"] || "" }}. - Ensure OpenAI Chat Engine is connected as the language model for Review Insight Agent.
- Credential Required: Connect your openAiApi credentials in OpenAI Chat Engine (credentials are managed on the model node, not the agent).
- Confirm Review Insight Agent outputs to Flatten AI Output for JSON parsing and field flattening.
⚠️ Common Pitfall: If the AI returns non-JSON text, Flatten AI Output will set parse_error and no KPIs will be appended. Keep the agent prompt unchanged to enforce strict JSON output.
Step 5: Configure Outputs and CRM Update
Append KPIs to the output sheet and update Salesforce when a CRM ID is present.
- In Append Sheet Record, keep Operation set to
appendand verify the columns map to AI fields like{{ $json.sentiment_positive }}and{{ $json.recent_trend_summary }}. - Configure Check CRM ID with the condition
{{ $('Fetch Sheet Rows').item.json.salesforce_account_id }}notEquals"". - In Trim Trend Summary, keep the truncation logic to enforce the 255-character limit before CRM updates.
- Configure Upsert CRM Account with Resource set to
accountand Operation set toupsert. - Set Name to
{{ $('HTML Text Cleaner').item.json.entity_slug }}and External ID Value to{{ $('Fetch Sheet Rows').item.json.salesforce_account_id }}. - Under Additional Fields, map feedback__c to
{{ $json.recent_trend_summary }}. - Credential Required: Connect your salesforceOAuth2Api credentials.
- Confirm the execution loop: Append Sheet Record → Check CRM ID → Trim Trend Summary → Upsert CRM Account → Batch Items Iterator to continue processing remaining rows.
⚠️ Common Pitfall: If salesforce_account_id is empty, the flow will skip Upsert CRM Account and return to Batch Items Iterator. Make sure the ID is present for records that should be updated.
Step 6: Test and Activate Your Workflow
Run a manual test to validate scraping, AI parsing, sheet updates, and CRM upserts before activating.
- Click Execute Workflow to run Manual Launch Trigger.
- Verify Fetch Sheet Rows loads URLs and Decodo Scrape Task retrieves HTML content.
- Check Flatten AI Output for parsed KPI fields (no
parse_error). - Confirm Append Sheet Record writes a new row to the output sheet.
- If a CRM ID exists, confirm Upsert CRM Account updates the record with
feedback__c. - When satisfied, toggle the workflow to Active for production use.
Common Gotchas
- Salesforce credentials can expire or need specific permissions. If things break, check the connected app/OAuth permissions and the n8n credential test 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.
- Decodo and OpenAI both enforce limits, and failures can look like “random” errors. If you see intermittent scrape or AI timeouts, slow the batch size and add a retry strategy.
Frequently Asked Questions
About an hour if your API keys and Salesforce access are ready.
No. You’ll connect accounts, paste in API keys, and map a couple of fields in n8n.
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 Decodo scraping costs and OpenAI API usage (usually a few cents per batch, depending on volume).
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 it’s a common tweak. You’ll map a different target field in the Salesforce upsert step, and you may adjust the “Trim Trend Summary” logic so the text fits the field length. Teams often customize the AI prompt to match brand language, add a “risk flag” for low ratings, or split output into separate fields for positives and negatives.
Usually it’s expired OAuth credentials or missing permissions to update Accounts. Reconnect Salesforce in n8n, confirm the user can edit the custom Account field, and double-check that you’re only updating existing Accounts (this workflow does not create new ones). If it fails only on some rows, it can also be a bad Account ID coming from Google Sheets.
On n8n Cloud, it depends on your execution limits and how you batch the sheet rows, but dozens to a few hundred reviews per run is normal. If you self-host, you’re mostly limited by your server and the rate limits of Decodo and OpenAI.
Often, yes, because this workflow needs batching, text cleanup, and AI steps that benefit from more flexible logic. n8n also gives you a self-host option, which can matter when you want higher volume without paying per task. Zapier or Make can still work if you keep it simple, like “new row in Sheets → send a Slack alert.” If you’re not sure, the fastest path is to describe your volume and goal and Talk to an automation expert.
Once this is running, your CRM stops being a place where feedback goes to die. The workflow keeps the sentiment fresh, and your team finally sees the themes without digging.
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.
