This post is part of Blue Seedling’s Hackathon Series, where we’ll share the good, the bad, and the great from our 2023 Generative AI Hackathon. The goal? Solve everyday problems at Blue Seedling using the power of AI.
A couple of months ago, we held our annual Blue Seedling team offsite in NYC, which was an absolute blast… for the most part.
In our post-offsite survey, most team members ranked our half-day generative AI hackathon as the worst experience of our 3-day offsite. 😱 You can read the gory details here. The TL;DR – the tools aren’t good, basic capabilities are missing, and the gap between expectations and reality is painfully wide.
We started the hackathon confident that after four hours, we would have working prototypes (MVPs) for all projects. No team did, and the process was extremely manual and frustrating.
But then… we rallied. True to Blue Seedling’s values, we problem-solved, continued working on our projects, and presented our updated MVPs two weeks later. Let me tell you – we were all blown away by the results.
Watch this space for a post by each team, sharing their rags to riches GenAI story. Here’s the story from Tim Jung and me.
Original idea: Create standalone ebooks with Blue Seedling’s content + GenAI
Blue Seedling has been creating original, high-quality, mostly evergreen content for 8 years. We have hundreds of posts about B2B marketing topics – 29 posts about email marketing, 45 posts about content marketing, you get it.
Our idea: Let’s have GenAI create standalone ebooks based on our entire library of content. For example, create “The Ultimate Guide to Enterprise B2B Email Marketing” based on the 45 relevant posts on our blog.
Spoiler alert: This failed miserably. 😭
Here’s why:
- Web content ingestion at scale is borderline impossible. We spent the majority of the hackathon’s four hours trying to get ChatGPT / Claude / Bard to “read” the hundreds of posts on Blue Seedling’s blog. We tried plugins, dropping a list of URLs, and other manual and cumbersome hacks. Ultimately, we gave up, and fed the tools only 18 of our best email marketing posts (we still don’t know of a GenAI tool capable of scanning all pages of a website. Please share recommendations!).
- GenAI long-form content creation is bad. My current favorite GenAI content use case is short-form co-creation: Dropping a rough draft of short-form content — like a LinkedIn post or a blog post — and working alongside ChatGPT (or Claude) to get to the final version.
But here, our task was different: We were creating content from scratch (based on our blog posts), and we wanted long-form content: a 10-15 page ebook, not a 500-word blog post. These factors combined led to dozens of back and forth prompts about the ebook’s outline and tone of voice.
Ultimately, we felt that getting the ebook to where it needed to be would be faster – and less frustrating – if we just wrote it ourselves using ChatGPT as an assistant / intern, not a co-creator.
The pivot: Using GenAI to create ultra-short-form content
Our a-ha moment arrived when we decided to pivot on one of the factors mentioned above – content length. We switched from long to (extremely) short-form content:
Can we use GenAI to tease out the hidden gems in our content mine?
Turns out, we can.
Use Case #1: Tips / Micro-advice
We asked ChatGPT to produce a series of short tips based on all our email marketing content, and we packaged those tips to be delivered daily. Every day, a reader gets one tip via email… or via a real-life “Page a Day” calendar!
For a full MVP, we found an Excel template for Page-a-Day calendars, and produced the first 10 days of January, using ChatGPT for formatting.
We added a Stable Doodle image as the cover image…
… and our humble MVP came to life:
Use Case #2: Short form content as an art form
Lastly, for fun, we asked ChatGPT to come up with haiku based on our email marketing content. We loved them.
ChatGPT’s capabilities really shine here. It’s so much easier for ChatGPT to adhere to the strict haiku format.
Takeaways for B2B Marketers creating content with GenAI
If you’re like us and you’re lucky enough to have a wealth of content at your disposal, here’s what we recommend:
- The secret to happiness is low expectations: Currently, most GenAI tools promise the world, but end up being a disappointment. When creating content with GenAI, calibrate your expectations for a realistic outlook on timeline and capabilities.
- Start small, test, pivot: Instead of aiming for a grand slam with a complex project like an ebook, consider starting with smaller, more manageable tasks. You’ll quickly learn what the tool can and can’t do, allowing you to adjust your strategy.
- Short-form magic: GenAI excels in generating short-form content, from daily tips to haiku. It’s a fresh way to repurpose existing content and deliver value in a bite-sized format. Not sure how a haiku fits into your marketing strategy? Be creative, or ask ChatGPT. 😉
The bottom line
Our content creation journey with GenAI was a rollercoaster, but it taught us valuable lessons. If you’re sitting on a treasure trove of content, GenAI can help you unearth bite-sized gems, even if it’s not yet cut out for heavy lifting, like ebook creation. Start small, be realistic, and you might just find a new way to engage your audience.
More from our Hackathon Series
- Buyer’s Guide in 5 min: Hacking long-form content creation using GenAI
- Automating a content calendar (and filling it with meaningful blog post topics) using ChatGPT
- Should you let ChatGPT write your newsletter?
