I'm building an online pet memorial with a song, video, and letters. Is that too much?

While reviewing the PawsLullaby landing page before launch, I noticed that the copy kept describing it as a “private pet memorial.”

That was accurate. It was also hiding the product.

If I removed the PawsLullaby name, the same sentence could describe almost any memorial website with a privacy setting. I had used a trust feature as the main reason to choose the product.

So I went back through the products I had researched. Most of them handled one part of remembrance well: a page for photos and an obituary, an AI memorial song, a slideshow, or some kind of chatbot or letter experience.

I could understand why. Each one is easy to explain. “Upload photos and create a page” is much cleaner than what I was building.

But it also creates a strange experience for the owner.

They might tell one tool about the morning walks, tell another tool what the dog was like, upload the same photos somewhere else, and then start again if they want to write a letter. The outputs can all look polished while feeling as if they belong to different pets.

That became the product decision behind PawsLullaby: one pet's real story should shape the whole Memory.

Generating four separate outputs is not the hard part. The hard part is making them feel as if they came from the same pet's life, without asking the owner to fill in one enormous form.

Each Memory starts with one structured story: the pet's name and dates, personality, routines, specific memories, the tone the owner wants, and the real photos and clips the owner chooses to upload.

That same source then drives different parts of the product:

• the page holds the story and finished pieces;
• the song uses the details and emotional direction;
• the video is made from the owner's real media rather than invented images;
• the letters carry the same voice and can use themes from a reply later.

I ended up separating the flow into stages:

one story → free preview → full media generation → paced letters

Before payment, PawsLullaby creates a lighter preview: the tribute line, letter opening, song direction, and video direction. The full personalized song and HD tribute video sit behind the unlock. People should be able to see whether the tone feels right before paying. It also stops the product from paying for full song and video generation every time an anonymous visitor tries the form.

The letters created a different kind of engineering problem.

They are not scheduled from the day someone creates an account. They are scheduled from the pet's actual passing date.

That distinction sounds small until you consider someone whose pet died two years ago. Sending them language written for the first 72 hours would be careless. The scheduling logic has to place them at the right point in the letter arc rather than mechanically starting at the beginning.

Replies also cannot be treated as ordinary chat history. A reply goes through a safety check. If it passes, its themes and emotional context can help shape a later letter. If it is held by the safety gate, it is not carried forward into generation.

I also split the private owner view from the page someone else can visit. The owner can see materials, replies, letter controls, and the complete dashboard. A share link receives only the parts the owner has chosen to make visible. Replies never appear there, and even delivered letters stay private unless the owner shares them individually.

This is still an AI-assisted product, but I do not want it to behave like an unlimited “talk to your pet again” chatbot. There is no fake pet voice, no invented photograph, and no claim that a generated letter is real contact or proof of an afterlife. The correspondence is paced, finite, and can be paused.

Those boundaries make the demo less spectacular. They also make the product feel safer to build.

Privacy is still built into the product, but it is no longer the headline. The actual product is the complete Memory: the page holds everything together, the song is the easiest entry point, the video makes the idea tangible, and the letters give someone a reason to return after the first visit.

That leaves a positioning question I have not solved.

If I lead with the whole Memory, the product may sound broad: page, song, video, and letters. If I lead with only the song, it is easier to understand, but it hides the system that makes the song feel connected to everything else.

I wrote the user-facing version of the idea here: How to create an online pet memorial.

If you landed on the product today, which would help you understand it faster: “an online pet memorial” or “a personalized pet memorial song”?