Building a Reddit listening tool with zero coding |
|
In this week's newsletter, I'm sharing a Reddit automation that our team built that completely changed how we approach Reddit marketing. Because I'll be honest... We struggled with it for a long time. When Reddit suddenly became an important visibility channel due to AI search, we had no idea how to make it work as an “influence” layer. But after months of trial and error, we finally found a workflow that actually worked. And the biggest driver for that was a Reddit listening tool our SEO manager, Kashish, built. In today's issue, I sat down with her to get the full scoop on how she built it, and the thinking behind it 👇 |
|
🛠️ Foundation behind the tool |
|
The entire workflow runs inside Gumloop, a no-code automation platform. Kashish connected a Reddit search node, an AI summarizer, a categorizer, a filter, and two output destinations without writing any custom code. The workflow sends everything to two places. ✅ Google Sheets, where it becomes useful for research, content planning, and topic mining. ✅ And Slack, where the team can quickly jump into relevant conversations. |
|
Step 1: Search Reddit The first node searches Reddit using a predefined list of keywords like email marketing, email automation, onboarding, and more. You can choose how many posts to collect and the time frame whether that's the past day, week, or month. For our workflow, we run it weekly and collect the 15 most relevant posts with their title, URL, full post content, and comments. |
|
Step 2: Summarize each thread Next, every Reddit thread goes through an AI summarizer agent. This reads the entire post along with its comments and generates a concise summary rather than just focusing on the title. This is because a title rarely tells the full story. For example, a post called "Quick question" could easily contain detailed buying intent buried inside the discussion. |
|
Step 3: Categorize the conversationEach summarized thread is passed to the next agent, which is a categorizer that classifies the queries into six categories: - Positive
- Commercial
- Promotional
- Informational
- Negative
- Not Relevant
|
|
Step 4: Filter the queries Every thread marked as Not Relevant gets removed by the filtering agent. We intentionally do the filtering after categorization because if we filtered during the Reddit search stage, we'd likely be filtering too early. A relevant discussion can easily appear in a slightly different language than our keywords. Also, by this stage, the AI has already read and summarized the entire conversation and determined it's seen relevancy. |
|
Step 5: Save everything to Google SheetsEvery thread that passes the filter gets logged into Google Sheets. Each row includes: - Post title
- Summary
- Full content
- Comments
- URL
- Category
Over time, it becomes a searchable database of customer pain points that the team can use for content planning, topical research, and product feedback. |
|
Step 6: Send it to SlackThe same threads are also pushed into a dedicated Slack channel. Each message includes the title, summary, content, and the Reddit link. Someone can spot a relevant discussion and jump into it via the shared link. |
|
AI news: What's happening right now? | - Dropbox just deepened its integration with OpenAI, letting users organize files, create shareable links, and run multi-step workflows directly inside ChatGPT, all pulling from content already stored in Dropbox. For GTM teams, this means AI-generated campaign briefs, launch assets, and messaging docs can now be created, saved, and shared without ever leaving the tools your team already lives in.
- AllyHub just launched a browser-native AI copilot that runs research, content creation, and repetitive web workflows directly inside the browser. This means if you were to do competitor research, content drafts, and data analysis, you build the workflow once and it keeps improving automatically.
- Edge226 acquired AnyClip, an AI video intelligence platform that analyzes and understands video content at scale, combining it with Edge226's existing cross-channel performance infrastructure across CTV, in-app, and mobile. For GTM teams this means contextual video targeting and performance measurement are getting closer to living in one place rather than being stitched together across tools.
|
|
|
| That's all, folks! I'll see you soon with more tips and ideas. If there is something we can do better for you, please let me know by replying to this email. |
|
Until next time,
Aquib
CEO, Mailmodo |
|
|
|
