How I Used AI for a Value Proposition and Image Selection

Yesterday I was working on a value proposition for our maintenance programs. We have excellent programs, but we need to shout out how we are different and better (which is our approach to how we do this for clients).

The messaging also needs to be interesting to grab the attention of the page’s visitors.

Our site maintenance programs have hours that don’t expire like many maintenance programs do so I was thinking what are some interesting ways I can talk about this?

I thought of the way a bank works, or how airlines store up frequent flyer miles.

I fed my challenge into ChatGPT and got 10 single sentence messages centered on maintenance hours that rollover.

The sentences were a little clumsy and not the voice I wanted but one statement stood out to me which was “bankable maintenance hours.” I had already thought about the bank comparison, but I really liked that statement.

I made a few more notes from the ChatGPT suggestions about other value points with rollover hours such as flexibility, peace of mind, and budget predictability. I felt I had enough to continue on my own.

I then turned to image selection, and this can always be challenging when talking about services that don’t easily translate to clever imagery. I wanted some concepts to think about.

I asked ChatGPT for image ideas for website maintenance programs and it came up with some clever ideas such as hands shaping a clay pot, a jeweler putting a watch together and a guitar luthier creating a guitar.

I wasn’t really happy with those ideas as they seemed too generic so I dug deeper feeding ChatGPT information about how competitive maintenance hours disappear for their clients at the end of the month.

This got more interesting with image ideas such as a person walking into fog, physical things breaking apart and smoke.

The person walking in fog images just seemed a bit dark and mysterious and things breaking up was too abstract so I started looking for artistic images of smoke and found several that might work.

I was honing in on ideas now and thought disappearing maintenance hours was something people can wrap their heads around.

What I did was use AI for some “conversation starters” and then some concepts on imagery that I fine-tuned. Most of what I got back I didn’t use but I did pull some clever ideas out of the process and kept going which saved me some brainstorming time. But I still came back to my process, I was just armed with ideas earlier in the exercise.

Feeding AI agents with as much information as you can and trying different angles to your challenges can provide you deeper and more focused feedback. Just don’t make the mistake of having AI attempt to solve all your challenges or completely do the writing for you. It’s guaranteed not to be the path your brain would take — or your company’s voice.

Here’s the value proposition copy so far for the landing page. It’s still a work in progress. What do you think?

Maintenance Hours Going Up in Smoke?

Our Plans Never Disappear

Start With as Little as an Hour a Month

Bankable Hours That Rollover Automatically

So go deep with your AI prompts and pull ideas where you can but you’ll still need that number 2 pencil for the mojo!

Summary

AI works best as a creative catalyst—not a replacement for human thinking. In this article, we explain how AI can accelerate brainstorming by generating initial language and visual concepts, while the real value still comes from human judgment, refinement, and brand voice.

By using AI to spark ideas around value propositions and imagery—and then applying experience and persistence to shape them—the result is stronger, more original messaging. The takeaway: use AI to start the conversation, not finish it.

FAQ

Is AI useful for creative work like branding and messaging?
Yes. AI is great for idea generation, pattern recognition, and exploring angles you might not think of right away—but it shouldn’t be trusted to do the final creative work.

What’s the biggest mistake people make when using AI for marketing content?
Letting AI fully write the copy or define the strategy, which often results in generic language and a loss of brand voice.

How can AI save time without hurting quality?
By accelerating early-stage brainstorming—headlines, metaphors, positioning ideas, and visual concepts—so you can spend more time refining and improving.

How detailed should AI prompts be?
The more context and constraints you provide, the more focused and useful the output will be. Shallow prompts lead to shallow ideas.

What still requires human input?
Judgment, taste, brand voice, strategic nuance, and knowing what not to use. That’s where the real value lives.