The Photobook Problem We Didn't Expect

The moment we realised the real problem wasn't collecting photos

Early on, everyone told us the hard part would be getting guests to upload their pictures. Make it frictionless, they said. No app, no login, no friction. We got that bit right. A QR code at the event, guests scan, browser upload, done. We had that working before we left alpha.

What we didn't anticipate was what happened after the event ended. Hosts would message us weeks later saying something like: 'I've got 600 brilliant photos. I want to print something but I can't decide which ones to include.' One user spent five evenings going through photos one by one, comparing angles, checking if someone's eyes were closed. She abandoned the project halfway through.

That's when it clicked. The bottleneck wasn't upload. It was the gap between 'we have all the photos' and 'we have a finished keepsake we're proud of.' Most people don't want to be photo editors. They just want the memory.

Why automation meant something different to us

We could have built a template tool. Let hosts choose layouts, drag photos around, pick fonts. That's what every other service does. But that's still asking people to do the work. We thought about what would actually reduce friction, and the answer was: don't ask them to choose at all.

For Occasion tier users, the photobook generation works like this. Host opens the gallery. Hits 'Create Photobook.' Selects a layout and hardcover style. The system runs through every photo in the album, applies basic quality checks (exposure, blur), and builds a book automatically. No manual curation. No 'which twenty photos should I use.' All of them.

The first time we tested this internally, I watched someone order a 150-page wedding photobook in ninety seconds. Her only input was 'yes, I want this.' That felt right.

The details that made the difference

Automation sounds simple until you start building it. We learned fast that there's a difference between 'automated' and 'automated well.'

The system had to handle photos at wildly different qualities. At an indoor birthday party, half the photos are blurry or underexposed. At an outdoor wedding, someone's pointing a phone camera at the sun. We built in detection for obvious failures (completely dark, completely blown out, etc.), but we kept most photos anyway. People often had favourites that weren't technically 'perfect.'

We also had to decide: portrait orientation or landscape? Mixed? We settled on letting hosts pick a layout style that then intelligently arranges photos. Some books end up 140 pages, some 80. That varies based on the actual event photos, not based on arbitrary slots we force them into.

The thing that genuinely surprised us was video uploads. Some Occasion tier hosts had mixed photos and videos in their gallery. Early on, the photobook just ignored videos. After a few customer messages, we realised people wanted those memories included too. So we built it so the system includes a QR code in the printed book that links to video clips. Physical book, digital video layer. Sounds gimmicky when you say it out loud. Feels genius when you actually use it.

What we still get wrong

Not every automated photobook comes out perfect. We get messages occasionally from hosts saying 'there's a photo of my feet in the middle and I don't know why it was included.' Fair point. No detection system is 100 percent accurate. And sometimes the order feels random to the person who was there.

We've learned that transparency matters here. Hosts can preview the book before it prints. If they hate the auto-order, they can manually reorder pages, remove photos, add captions. The automation is a starting point, not a prison.

What we're still refining is how to handle galleries with extreme quantities. We had one corporate event with over 2,000 photos. The automated book was 280 pages. Beautiful, technically. But maybe too much for one physical object. We're thinking about how to offer hosts better guidance on that, without making the process complex again.

The reason this matters beyond the feature

Here's what I think matters: people collect photos now because cameras are everywhere. But they rarely finish that collection into something physical. There's a graveyard of photo folders and cloud albums that never become objects.

Poolr's job was always to make collecting photos at events dead simple. The photobook automation extends that. It completes the journey from 'we all took pictures' to 'we have something we can hold and remember this by.' That's the whole point.

When Sophie got that printed hen-weekend book, she didn't think 'wow, the automation was impressive.' She thought 'we have this memory now.' That's the only metric that mattered.