What AI Taught Me About Pinterest (And It Wasn't What I Expected)

Welcome to the Journal: a place where I document experiments, observations, and lessons learned while building this blog.

Disclaimer: This isn't a scientific study.
It's simply one creator's experience with a small Pinterest account over the past few months. Your results may be completely different.

A bit of background on this blog. I started the site with a fitness and wellness focus. I enjoy inspiring people to exercise and take care of themselves. Getting that inspiration out took several forms as I tried different social media platforms to see which would be the most manageable for me and my lifestyle (I am not a full-time blogger).

Out of the different social media channels, I found that Pinterest was relatively easy to keep up with and post to regularly. My page was growing slowly but steadily. Most of the pins I put out were created by me using Canva or simply from my own photos.

Fitness pins on Pinterest aren't exactly among the platform's trending content, so out of curiosity I decided to experiment with AI-generated nail inspiration instead. Nail inspiration is some of the best-performing content on Pinterest, and I wanted to see what would happen.

The results surprised me and raised more questions than answers.

What happened?

I already had some outfit pins that had been doing well. They got maybe a couple of thousand impressions, out of them maybe a couple of dozen outbound clicks to my blog articles. I switched entirely to posting AI-generated nail inspiration. Now I love nail art and have dabbled in it quite a bit, but I’ll be honest when I say that I haven’t painted my nails once in the last 2 years. I would have loved to mix in some actual nail art just to compare the results, but I simply don't have the time! :D

I kept the same posting schedule throughout the experiment, with one pin scheduled every other day. The images were clearly labelled as AI-generated.

Within about a week, the nail inspiration pins had completely overtaken my other content. Pin impressions (the number of times a pin appears on someone's screen) grew dramatically. The pins also received far more saves and likes than my fitness content.

Despite the reach, however, there was almost no follower growth. The saves and likes, even though they outperformed my fitness content, were still surprisingly small compared to the impression numbers.

To give you an idea, over a 30-day period my pins received around 380,000 impressions, but my engaged audience (the number of people who interacted with my pins) was only 11,000. During that same period, people saved my pins around 1,700 times. My best-performing pin reached 183,000 impressions but received only 31 likes.

The numbers confused me, to say the least.

I also discovered something else while scrolling through the suggested pins underneath my own AI-generated images: a plethora of almost identical aesthetic nail designs. Pinterest is flooded with the same style of AI-generated content. A simple search produces what feels like an endless sea of very similar-looking images.

The AI question

Trying to explain these results to myself certainly raised more questions I’m not sure I’ve figured out the answers to.

Are people comfortable with AI if the result is useful? - My own Pinterest experience suggests that usefulness may matter more than I expected. The pins were clearly labelled as AI-generated, yet people still clicked on them in large numbers. Whether they saved them because they liked the designs or simply because they were eye-catching, I can't really tell.

Does transparency matter? - I'm glad Pinterest labels AI-generated images. Even if someone doesn't mind AI, I think it's fair that they know what they're looking at. That transparency lets people decide for themselves.

Can AI create attention but not trust? - This is probably the biggest question I came away with. The impressions exploded, but my follower count barely moved. That made me wonder whether AI images are excellent at grabbing attention but much less effective at helping people connect with the creator behind them.

Is AI just another creative tool? - This is the question I'm still wrestling with. I don't see AI as inherently good or bad. I used it because I was curious and wanted to experiment. But I also don't want every piece of content I create to feel interchangeable with thousands of other images. Somewhere between those two extremes is probably where I'd like to be.

What I'm taking away

This experiment hasn't convinced me that AI is the future of content. It has convinced me that platforms, audiences and success metrics are changing faster than I expected.

I'm still interested in creating fitness and wellness content, but this experience has made me rethink how people discover it. Looking at Pinterest now, it's hard to ignore just how many visually similar AI-generated images there are. It makes me wonder whether creating something genuinely unique has become harder over the last few years but also more valuable than ever.

I'm going to keep experimenting and documenting what I learn. If you're a creator who's noticed similar changes (or completely different ones), I'd love to hear your perspective.

Update (coming soon):
I'm continuing to test different types of Pinterest content. If anything changes significantly, I'll come back and update this article.