Creating a FAQ list from your HubSpot conversations can help you address customer needs efficiently, reduce repetitive questions, and improve your support resources.
This guide outlines a step-by-step approach, from exporting your chat data to compiling actionable FAQs.
Why FAQs Matter
- FAQs streamline customer support and empower users to find answers quickly.
- HubSpot chat conversations are a goldmine for identifying common customer questions and concerns.
Step 1: Retrieve your HubSpot conversations data
To fetch your conversations data from HubSpot, you can make a custom integration to call the HubSpot API or use Conversations Export to export the conversations for you.
- To use the HubSpot API
- To use Conversations Export
- Connect the Conversations Export integration: Follow the getting started guide here to open the app in your HubSpot portal.
- Click Add Export and select your workspace and channel: You can choose from workspaces like Inbox, Help Desk and channels like Live Chat, Facebook Messenger, etc.
- Check anonymize conversations: This internally (no 3rd party data access) does an initial pass of redacting personally identifiable information like names, phone numbers, addresses, etc.
Warning: some PII may remain due to data structures. Always check your conversations data for PII before uploading to any AI or analytics tools.
- Optionally choose a date range: Set this if you want to restrict the data to a certain period of time.
- Download the File: Some larger exports can take considerable time. You can close the app and receive an email with a link when the download is ready.
Step 2: Preparing the Data for Analysis
- Open the Exported File: Use spreadsheet software (Excel, Google Sheets) to view your chat data.
- Review Data Structure: Identify columns such as timestamp, agent, and conversation content.
- Check for personally identifiable information: Remove any references to private information like names, phone numbers, addresses, etc. Anonymize conversations will do an initial pass but due to varying data formats, may not catch every reference.
Step 3: Use a prompt to process your conversations with AI
You can use a free Perplexity.ai account to analyze this data for you. Remember to double check for personally identifiable information before uploading.
- Prompt:
Act as a customer support analyst. You are provided with a dataset of HubSpot live chat conversation transcripts exported from our CRM.
Your tasks are:
- Analyze all chat transcripts and identify the most frequently asked questions by users.
- For each question, group similar questions together (even if phrased differently) to ensure accurate frequency counts.
- List the top 10 most frequently asked questions in order of occurrence.
- For each question, provide a brief summary of the common intent or issue behind it.
-
Output the results as a markdown table with columns:
-
Question
-
Frequency
-
Summary of Intent/Issue
-
Step 4: Drafting Your FAQ List
- Write Clear Answers: For each frequent question, draft concise, helpful responses.
- Use Customer Language: Phrase questions and answers using the same language your customers use in chats for clarity and relatability.
- Review and Refine: Edit for accuracy, tone, and completeness.
Step 5: Updating and Maintaining Your FAQ
- Regularly Re-Export Chats: periodically export your chats to keep your FAQ list current as new questions emerge.
- Solicit Feedback: Ask customers and support agents for suggestions on FAQ improvements.
You now have a comprehensive, user-driven FAQ!
By systematically exporting, reviewing, and analyzing your HubSpot conversations, you can build a dynamic FAQ resource that saves time for both your team and your customers, while continuously improving your support experience.