In the second episode of the Hustles Debunked series, the challenge was clear: could I create a daily newsletter that starts making money within 30 days? This time, the project was to build a "Nostalgia Newsletter" — a daily email written as if we were living 30 years in the past, revisiting the year 1993. Here’s a deep dive into the process, the tools, and the lessons learned from using ChatGPT and automation to bring this idea to life.
The Concept
The core idea was simple yet appealing: deliver a daily dose of nostalgia by highlighting key events from 1993. The newsletter would feel like stepping back in time, offering subscribers a fresh yet familiar experience. But the big question remained: Can we monetize it quickly?
Early Challenges with Design
While ChatGPT excels at crafting text-based content, it struggles with visual design. My first attempt to use GPT for the email layout itself was disappointing. So I turned to HubSpot’s email builder, creating a clean, professional template. This allowed me to simply plug in generated content, focusing the AI’s work on writing rather than design.
This division of labor was essential—design was handled manually while ChatGPT took on content generation.
Automating Content Generation with a Tech Stack
Running a daily newsletter is no small task, especially with a full workload already in place. Automation became a necessity, and here’s how I put it together:
-
Triggering with Google Sheets and Zapier: I set up a Google Sheet where inputting a date would trigger a Zapier automation. This kickstarts the entire content creation and delivery pipeline.
-
Collecting Source Data: To ensure the newsletter was factually accurate, I first pulled historical facts from a website called Best of Date for each day’s events. Early attempts to rely solely on GPT’s knowledge led to inaccuracies, so grounding each newsletter in real data was a must.
-
Using ChatGPT for Writing, But With Guidance: Next, I fed the collected facts into carefully crafted GPT prompts to generate different newsletter sections: an editorial intro, main story, second story, and titles. Each part was handled with a dedicated prompt in order to structure the email properly.
Since Zapier’s native OpenAI integration lacked GPT-4 access, I used Human Loop to bridge between Zapier and GPT-4. This setup let me test and tweak prompts without having to rebuild the automation each time.
- Avoiding Repetition: A tricky issue was GPT occasionally picking the same news item for both the main and secondary stories. To address this, I provided previous days’ news as context in prompts to reduce repetition and keep the content fresh.
Handling Images: A Manual Touch in an Automated Process
Visuals are critical to newsletter engagement. Ideally, I wanted to generate images automatically within Zapier using DALL·E. However, the AI-generated images were lackluster, so I switched to using MidJourney for better quality artwork.
The workflow for images looked like this:
- GPT wrote a prompt describing the image needed for each story.
- I manually input those prompts into MidJourney to generate the images.
- Once approved, I uploaded the images and replaced the image URLs in the template.
Although this step required manual effort, I felt it was worth the time because images greatly enhance the newsletter’s appeal. Unfortunately, there is currently no API to automate MidJourney image generation, so if you know one, reaching out would be appreciated.
Final Assembly and Delivery
After receiving all written content and images:
- The Zapier automation replaced the placeholder text and image URLs in the saved HTML.
- A test email was sent to me to ensure formatting and content looked polished.
- Then, the final HTML code was uploaded back to HubSpot.
- HubSpot’s email workflows took over, scheduling each email to be sent automatically at 5 a.m. every day.
Because the newsletter is based on 1993 news, all future emails could be scheduled in advance, meaning new subscribers would automatically get subsequent emails without any additional work.
Reflections on the Hustle
This experiment highlighted both the tremendous potential and current limitations of AI for newsletter creation:
- GPT-4 is excellent at text generation but requires well-crafted prompts and reliable data inputs.
- Automations can drastically reduce daily workloads, but some manual effort—especially for design and images—still enhances quality.
- Using a tech stack involving Google Sheets, Zapier, Human Loop, and HubSpot created a streamlined flow that could be customized and scaled.
Most importantly, the Hustles Debunked mission means I am testing these ideas in real time, with no guarantee of success. Whether this newsletter becomes a viable business or not, it’s a valuable learning experience in combining AI with automation to create engaging content at scale.
If you are curious to see whether this hustle pays off before the 30-day mark, following this series is the best way to stay updated.
To wrap up: Crafting a daily newsletter using ChatGPT is feasible but requires a hybrid approach—human creativity alongside AI efficiency. This Nostalgia Newsletter project is a testament to that balance and a glimpse into the future of automated content creation.
Stay tuned for the next update in the Hustles Debunked series!
—————————————————-
Are you tired of the 9-to-5 grind? Unlock the secrets to online income generation. GetIncomeNow.com is your roadmap to financial freedom. We reveal proven methods, insider tips, and cutting-edge strategies to help you achieve your income goals. Start your journey to financial independence today!