Our approach to getting into AI is to simply dive in by experimenting. This blog post is the first in a series where we will give you a glimpse into our initial experiences with AI and share department-specific insights.

We’re excited to present this first post, spotlighting the AI journey of our Marketing Department. Our aim is to spark your interest in AI and provide valuable insights on how you can incorporate it into your own company.

Experiment 1: AI-Generated Images for Blog Posts

We’re all about authenticity, which is why we steer clear of stock images. Our company and content are one-of-a-kind, and our images should also reflect that uniqueness.

Our goal was to craft a consistent and easily replicable style for our blog post images, capturing the distinctiveness of each post. We concentrated on making a set of rules to keep our images looking sleek and uniform. We leveraged AI tools like DALL-E and Midjourney, which are designed for generating and modifying images. What did we think of the results?

The first platform we experimented with back in September was DALL-E. To be frank, after testing various prompts, the style of the images didn’t quite align with our expectations. Here is what it produced:

Following a friend’s recommendation, we gave Midjourney a try, and we genuinely liked the community vibe it offered. It operates within a Discord group, where every chat message transforms into a prompt. It’s a dynamic, interactive space where you can experiment with ideas and draw inspiration from the collective creativity of others.

In fact, the featured image at the top of this blog post was created on Midjourney using the following prompt:

“Create a hyperrealist, futuristic image of a Latin woman in her thirties working remotely from her laptop. She is engaged with AI platforms for marketing. The perspective is from the side but the glow from her computer screen is visible. The scene should incorporate floating elements of marketing. Aspect ratio: 16:9.”

Our initial experience with DALL-E yielded results that were a bit cartoonish. But we stuck with it, trying out various platforms and prompts. Midjourney turned out to be a game-changer for us. It boosted our skills, and soon we were creating images that reflected our brand. It left us feeling prepared for another attempt with DALL-E and other platforms because we better understood how to leverage their capabilities and explore creative possibilities.

Experiment 2: Enhancing Video Content

Our chairman, Rob Levin, shines in public speaking but struggles with memorizing scripts. In a recent 30-second ad, it was obvious he was reading, hindering the crucial eye contact for audience connection. So, we got creative to find a solution.

Leveraging Captions AI’s eye contact feature seamlessly made Rob appear as if he’s looking right at the camera, effortlessly nailing the script without breaking a sweat. Once we found the right AI editing tool, we applied eye contact filters, reduced background noise, and added captions to boost overall engagement. How did it turn out? See for yourself:

Video before eye contact enhancement

Video after eye contact enhancement

This tool is revolutionary. Anyone can read from a script and appear to be speaking off the cuff, all thanks to AI creating a natural eye contact illusion. Our first try in September was decent, but it had some glitches. On our second attempt in November, the improvement was remarkable – almost perfect. It’s a clear testament to the swift evolution of AI tools!

Experiment 3: Using ChatGPT for HTML Code

In the marketing department, we constantly experiment to boost email open rates. Text-based emails have been successful in avoiding the feared promotions tab, but a challenge arose when sharing a tiered pricing table over email. HubSpot, our CRM, lacked table support for emails. With our goal of designing an effective, image-free HTML price table, we turned to ChatGPT to generate the required HTML, seamlessly integrating a table into our CRM’s email tool.

We started the process by using the free version of ChatGPT to generate the initial HTML code for our table. After that, we effortlessly embedded it into our HubSpot email template, tweaking and refining the design. To get the table just right, we provided more specific prompts to ChatGPT, giving clear instructions on elements like font, color, and spacing to align with our desired outcome.

The first HTML design from ChatGPT was solid, but it needed some fine-tuning. By providing more specific prompts and adopting a more conversational approach with the AI, we were able to refine the table to our exact needs. 

HTML Code pasted on HubSpot Email module and the final table.

After trying this out, we upgraded to the ChatGPT paid version, and wow, it could’ve saved us so much time! The image upload feature is a total game-changer, transforming images into code in a flash. It’s not just great for HTML; it’s a lifesaver for any custom code for our website and other projects. The best part? You don’t need to be an HTML guru. Just toss a prompt to the AI, and voilà, you get your code.

Our IT team has already harnessed it to whip up custom code, tailoring our recruitment platform (more on that in an upcoming post).

WBN’s AI Adventure Continues!

At WBN, we’re only getting started with the incredible possibilities AI has to offer. Despite working with a small-budget AI exploration, we’re constantly uncovering new ways to enhance our workflows. This AI journey is just a fraction of our broader mission to always stay ahead of the curve. We’re excited about the potential of AI and eager to share the journey with you. Across all departments, we’re diving into AI to revolutionize how we work. The adventure has just begun!

Stay tuned for our upcoming blog post as we delve into the transformative impact of AI on our IT and Project Management Department processes. It’s all about learning, growing, and innovating, and we’re excited to share it with you.

Ready to hire a WBN remote professional? Sign up for a free 15-minute consultation!