We were wrong about the AI pet photo market: it's not really about pets

When we started building Pet Imagination at Inithouse, a studio shipping a growing portfolio of products in parallel, our thesis was simple: pet owners want AI photos of their animals. Upload a photo, pick a style, get a portrait. Utility play.

We were wrong about who the customer actually is, and the correction changed how we build the product.

The hypothesis that didn't hold

Our initial framing: "pet photo tool." A utility. You have a dog, you want a cool picture, here you go.

But when we looked at how people actually use Pet Imagination (9 styles ranging from Renaissance and Watercolor to Astronaut and Final Boss), the pattern didn't match the hypothesis.

Users weren't saying "I want an AI photo of my dog." They were saying "I want to see Max as a superhero." Or "make my cat look like a wizard." The request wasn't about documentation. It was about play, identity, emotion.

What the style data told us

We offer 9 portrait styles. If our "photo utility" thesis was correct, you'd expect the realistic styles to dominate. Watercolor, maybe Sketch.

That's not what happened.

The most-used styles were the fantasy and costume ones: Renaissance, Sheriff, Wizard, Astronaut, Final Boss. The styles that transform the pet into something it obviously isn't. The styles where the owner is projecting personality onto the animal.

When we added Renaissance portrait and Superhero early on, engagement with the tool jumped. People weren't looking for accuracy. They were looking for expression.

The gift use case we almost missed

A significant chunk of our usage comes from the gift angle. "I want to give my partner a funny portrait of our cat for their birthday." That's not a pet product decision. That's an interpersonal gift decision. The pet is the vehicle, not the customer need.

This reframe matters for product decisions:

• Onboarding: we stopped leading with "upload your pet photo" and started with "what would your pet look like as a Renaissance noble?"
• Style priority: we invest more in expressive, shareable styles than photorealistic ones
• Pricing: free with no signup, because the share loop is worth more than a paywall. Every shared portrait functions as distribution

What this looks like in acquisition

Social sharing turned out to be our strongest channel. The typical journey: someone sees a friend's AI pet portrait on Instagram, thinks "I want one for my dog," searches something very specific like "turn my cat into a painting" or "AI pet portrait generator," and finds us.

SEO works, but the queries are emotional, not technical. Nobody searches "AI pet photo tool." They search for the transformation they want to see.

We observed a similar pattern at Magical Song, where users don't search for "AI song generator." They search for "custom birthday song for my wife" or "personalized wedding song." The underlying need is expression and gifting, not the technology. Across our portfolio at Inithouse, we keep finding that the emotional use case outperforms the technical framing.

The honest part

The AI pet portrait market is competitive and the underlying model is commodity. We don't have a moat in the AI layer. Our differentiation is UX, style variety, and the zero-friction flow (no signup, free, private, under 60 seconds). If someone builds a smoother experience tomorrow, we need to adapt fast.

What we'd do differently if starting over: skip the "photo tool" positioning entirely. Start from "emotional expression tool that happens to use pet photos as input." It would have saved us weeks of wrong onboarding copy and style prioritization.

The reframe that matters for builders

If you're building a consumer AI product, pay attention to the gap between what you think the product does (utility) and what users actually use it for (expression, identity, gifting). The functional description of your product might be technically correct and still completely miss the real value prop.

We measured this at Inithouse across several products in the portfolio, and the pattern holds: the emotional framing consistently outperforms the technical one in both conversion and organic discovery.

Try it yourself: petimagination.com