AUTHOR

GC Tsalamagkakis

How To Build An AI App That Counts Calories From Pictures With Bolt and ChatGPT

Zapier

ChatGPT

Reddit

Conducting Painstorming on Reddit with Zapier and ChatGPT 

Learn how to use Zapier and ChatGPT to identify customer pain points from Reddit and automatically generate a report that summarizes and prioritizes key issues. 

Every day, thousands of customers flock to Reddit to describe their biggest pain points. Those complaints represent potential product opportunities that are hiding in plain sight.

In this tutorial, you’re going to build an automatic painstorming system that helps you prioritize customers’ needs and maximize customer satisfaction:

Every breakthrough product begins with a problem that keeps someone awake at night. No matter how many patterns and insights you extrapolate from Excel spreadsheets, the best way to create products that people want is to start with the deep, persistent frustrations that customers personally wrestle with.

" framework flips traditional brainstorming on its head: instead of imagining solutions, you simply listen to what customers are complaining about. What problems are currently hurting your customers? This deceptively simple approach leads to finding problems that are worth solving.

When painstorming, you document specific complaints rather than abstract problems. Rather than noting "people struggle with productivity," you seek out raw, detailed frustrations like "I waste two hours every Monday morning copying bugs and suggestions from Discord/Slack into Asana." 

Reddit's ecosystem of niche communities makes it a goldmine for authentic pain points — users candidly share their struggles, workarounds, and moments of frustration. By regularly collecting and organizing posts of customers’ pain points from Reddit, you can have a steady stream of customer feedback to act upon every week without having to schedule and hold in-person interviews with customers. 

However, manually scouring Reddit for relevant complaints is both time-consuming and inefficient. That's where automation with Zapier and ChatGPT comes in. In the next section, you will set up a Zapier workflow to automatically fetch the top weekly posts from a target subreddit and funnel that data into a painstorming report that’s generated by ChatGPT.

Setting Up an Automated Workflow with Zapier

Creating a Zapier Trigger to Run a Zap Every Week

The first step to setting up a step-by-step, automated workflow with Zapier (also known as a “Zap”) is to add a 

. A trigger is an event that triggers a Zap. In this case, you will add a scheduled trigger that runs the Zap once every week.

In the Zapier dashboard, create a new Zap by clicking on the “+ Create” button, and then clicking on the “Zaps” option under the drop-down menu. 

Within the Zap editor, click on the node box labelled “Trigger.” In the pop-up modal, select the Zapier “Schedule” tool. 

Set the trigger event to “Every Week.”

Specify the “Day of the Week” as “Monday” and the “Time of Day” as 12:00 A.M. This means at the beginning of your work week, you can expect a painstorming report to be ready for analysis. 

To ensure that the trigger was configured correctly, click the “Test Trigger” button. If a test record appears, then the trigger works properly.

Fetching Top Weekly Posts From a Subreddit via Reddit’s JSON API

, raw JSON data of a subreddit’s top posts can be retrieved from the 

GET https://www.reddit.com/r/<subreddit>/top/.json

 endpoint. For example, to preview the top posts from the “iphone” subreddit in JSON format, enter 

https://www.reddit.com/r/iphone/top/.json

 into your browser’s address bar. By default, the top 25 posts posted within the past 24 hours are returned.

Although this data can easily be accessed in a browser, accessing this data in a Zap requires an additional step. When accessing this data from a third-party service like Zapier, Reddit’s API requires you to provide an additional “Authorization” header with a valid bearer access token. Otherwise, Reddit’s API will return a 403 Forbidden error.

To obtain a valid bearer access token for Reddit’s API via Basic Authentication, you must:

Obtaining a Client ID and Client Secret for Reddit’s API

To obtain a client ID and client secret, you must first register a new authorized application. Start by logging into your Reddit account and visit 

. Then, click on the “create another app…” button and fill out the required information. In this case, I’ve named the application “painstorming-zap” and set the redirect URL to 

Once you have successfully registered the authorized application, you can find the client ID and client secret under the application’s information (see labelled, bordered areas below). 

Creating Zapier Actions to Obtain an Access Token for Reddit’s API

Back inside the Zap editor, click on the node box labelled “Action.” In the pop-up modal, select the Zapier “Code” tool.

Then, set the “Action event” to “Run Python,” and press the “Continue” button.

In the “Code” textarea field, copy the Python code below and paste the code into this field, replacing 

 with the client ID obtained from Reddit and 

 with the client secret obtained from Reddit:

from base64 import b64encode

output = { 'encoded_client_id_secret': b64encode("<client-id>:<client-secret>".encode("utf-8")).decode("utf-8") }

To ensure that the action was configured correctly, click the “Test Step” button. If an “Encoded Client Id Secret” appears, then the action works properly.

Below the node box labelled “Code by Zapier,” press the “+” button to add a new action to the Zap. This next action will send a request that has an “Authorization” header with the base64-encoded string to the 

POST https://www.reddit.com/api/v1/access_token

In the pop-up modal, select the Zapier “Webhooks” tool.

Then, set the “Action event” to “POST,” and press the “Continue” button.

https://www.reddit.com/api/v1/access_token

 and the “Payload Type” as “form.” For the POST request’s payload data, add the following value sets:

Lastly, under “Headers,” add an “Authorization” header that’s set to 

, you can reference the base64-encoded string from the “Run Python” step by typing 

 and selecting “Run Python” → “Encoded Client Id Secret” from the opened pop-up.

To ensure that the action was configured correctly, click the “Test Step” button. If an “Access Token” appears, then the action works properly.

Creating a Zapier Action to Fetch Weekly Top Posts From a Reddit Subreddit

With a valid bearer access token, you have the necessary permissions to fetch weekly top posts from a Reddit subreddit. In the next webhook action, you will send a request to the 

GET https://oauth.reddit.com/r/<subreddit>/top/.json

 endpoint that has an “Authorization” header set to 

Below the node box labelled “Webhooks by Zapier,” press the “+” button to add a new action to the Zap. 

Then, set the “Action event” to “GET,” and press the “Continue” button.

For this tutorial, the Zap will generate a painstorming report for iPhones. Hence, this action will fetch weekly top posts from the “iphone” subreddit.

https://oauth.reddit.com/r/iphone/top/.json

. To ensure that Reddit returns the top weekly posts (and not the top posts from just the past 24 hours), add the 

 endpoint returns 25 posts. You can specify the number of posts returned by this endpoint by adding the query parameter 

https://www.reddit.com/r/<subreddit>/top/.json?limit=50

 will return 50 posts. For more query parameters, visit 

, you can reference the access token from the previous “POST” step by typing 

 and selecting “POST” → “Access Token” from the opened pop-up.

As before, to ensure that the action was configured correctly, click the “Test Step” button. If an “Access Token” appears, then the action works properly.

As you can see from the title of one of the top weekly comments, an iPhone user is complaining that their iPhone 15 Pro Max is burning up.

Generating a Painstorming Report From Subreddit's Top Weekly Comments with ChatGPT

Signing Up for an OpenAI (ChatGPT) API Key

To obtain an OpenAI API key, you must sign up for an OpenAI account and have billing set up for the account. Then, visit 

https://platform.openai.com/settings/organization/api-keys

 and click the “+ Create new secret key” button.

In the modal, provide a name for this API key. The name identifies the API key in the OpenAI developer dashboard, which allows you to know the exact usage metrics and costs associated with this API key.

Don’t worry too much about the cost since each invocation of this zap will cost less than $0.01 USD (using the GPT-4o mini model).

Once the API key has been created, copy the API key. 

Creating a Zapier Action to Generate a Painstorming Report with ChatGPT

Back in the Zap editor, below the last node box labeled “Webhooks by Zapier,” press the “+” button to add a new action to the Zap. 

In the pop-up modal, select the Zapier “ChatGPT” tool.

Then, set the “Action event” to “Conversation,” and give Zapier permission to access your OpenAI account by providing Zapier the OpenAI API key that you created in the previous step.

Once that’s done, press the “Continue” button.

Next, fill out the user message (the message that you would send as if you are interacting with the ChatGPT web interface) with the following prompt:

Below, delimited by triple backticks, is a collection of user feedback and complaints retrieved from the "<subreddit>" subreddit.

```json
<data>
```

Please analyze this data and generate a structured report following these steps:

Analysis Framework: 

1. First Pass - Surface Problems    
   * List explicit complaints and frustrations.
   * Note specific vocabulary and phrases users employ.
   * Identify common themes without interpretation.
2. Deep Dive - Root Causes
   * Examine what users were trying to accomplish when they encountered problems.
   * Identify unstated assumptions and expectations.    
   * Note workarounds and "jerry-rigged" solutions.
   * Look for signs of embarrassment, uncertainty, or resignation.
3. Pattern Recognition    
   * Group related pains by underlying cause.    
   * Identify recurring situations or contexts.    
   * Note demographic or experience-level patterns.    
   * Highlight emotional patterns in descriptions. 

Output Format: 

1. Executive Summary
   * Key patterns discovered.
   * Most significant pain points.
   * Notable audience segments affected. 
2. Detailed Findings
   * Primary pain categories with representative quotes.
   * Frequency and severity analysis.
   * Current coping mechanisms observed.
   * Hidden costs (time, money, emotional) to users.
3. Opportunity Analysis
   * Unmet needs revealed.
   * Gaps in existing solutions.
   * Potential solution spaces to explore.

Requirements: 
* Use direct quotes to support findings.
* Distinguish between explicit statements and inferred patterns.
* Note confidence levels for interpretations.
* Highlight unexpected or counter-intuitive findings.
* Include raw counts where relevant. 

Constraints:
* Focus on problems, not solutions.
* Maintain user perspective without technical interpretation 
* Avoid assuming causation without clear evidence.
* Note areas requiring additional investigation.

End your analysis with 3 key recommendations for further research based on the most promising or unclear patterns discovered. 

Remember: The goal is to understand the true nature of user pains, not just catalog complaints. Look for the story behind the feedback. 

Please begin your analysis.

, you can reference the content from every post from the previous “GET” step by typing 

 and selecting “GET” → “Data Children Data Selftext” from the opened pop-up.

If you want to format the data , then you will need to introduce an intermediate “Python Code” step between this step and the previous “Webhooks by Zapier” step that uses Python code to process the data.  

Then, specify the assistant name as “product researcher.” This can be anything, but ideally, you should name the assistant based on the role that you want ChatGPT to assume. 

Additionally, specify the assistant instructions as the following:

You are an experienced product researcher analyzing user feedback to identify deep-rooted pains and problems. Your goal is to uncover not just surface complaints, but underlying patterns that reveal true customer needs and opportunities. 

Once all of the necessary fields have been filled out, press the “Continue” button.

Learn

The newline Guide to Building Your First GraphQL Server with Node and TypeScript

Teach

Amelia Wattenberger

Author of Fullstack D3

Community

Tutorials on Chatgpt

How To Build An AI App That Counts Calories From Pictures With Bolt and ChatGPT

Conducting Painstorming on Reddit with Zapier and ChatGPT

This has been a really good investment!

Advance your career with newline Pro.

Email Newsletter

Popular Topics

Masterclasses

Tutorials

Fullstack React with TypeScript