On the surface, the growth problem looked familiar.
Get more of it.
Improve it.
Run them. Test offers.
We did all of those things.
But over time, the data began pointing toward a different strategy.
The promotion worked —
until it didn't
Promotions were extremely effective at producing short-term sales. Run a discount and orders jumped. Conversion increased. Revenue followed.
If we stopped the analysis there, the answer was obvious: run more promotions.
But we didn't stop there.
After promotions ended, sales frequently fell below their normal baseline. Customers weren't necessarily buying more; many were simply moving purchases they would have made later into the promotional window.
We were borrowing demand from the future.
That mattered even more for a premium brand. Repeated discounting risked training customers to wait for the next sale rather than giving them a reason to prefer Printique.
The experimentation data was pushing us away from promotion-led growth and toward something harder:
Build a better product experience.
But “better” wasn't
the same for everyone
Printique served two very different customers.
ProfessionalA professional photographer might care deeply about control: precise layouts, cropping, image placement, paper choices, color, and the ability to manipulate every detail of an album.
ConsumerA consumer making a book from a family vacation has a very different problem. They may have 500 photos on their phone and simply want to turn them into something beautiful without spending Saturday afternoon designing it.
We could build separate experiences for each audience. But that immediately creates another UX problem:
How do you know which experience to show someone when they first arrive?
The conventional answer is to ask.
Are you a professional photographer?
What are you creating?
How much control do you want?
We didn't want to put a survey between the customer and the product.
We already had a much better first action.
Give us your photos.
The photos became
the first CTA
That changed the experience.
Instead of asking customers to configure a product before they could see any value, we moved photo upload to the beginning of the journey.
Upload your photos first. Then let the product show you something.
That was especially important for consumers. People don't necessarily want to understand editing modes, templates, layout controls, or album architecture before they've begun.
They want to see their pictures in a book.
Once the photos are there, the experience can become progressively more sophisticated. A consumer can start from an automatically generated album. A professional can move toward greater control. We can worry about modes after we've given the customer something worth editing.
The first job of the experience is to get them to that moment.
AI made that first experience
dramatically better
Uploading hundreds of photos creates its own problem. The software now has a pile of images, but it doesn't understand them.
We used Google Vision to analyze the collection and extract information about its contents, including people and their prominence across the photo set.
Now 500 images didn't have to be treated as 500 unrelated files. The system could begin identifying patterns within the collection and use that information to make better decisions about how the album should be assembled.
AI helped turn an unstructured collection of photos into something the product could reason about.
The album builder could then generate a meaningful starting point immediately.
Instead of:
Upload
Configure
Design
Arrange
Maybe
see something good.
the experience became closer to:
Upload
See
your photos become a product.
That redesigned experience improved conversion by approximately 20%.
The data
led us there
The interesting part wasn't that one experiment failed and another succeeded.
Both could make conversion go up.
Promotionsincreased conversion by changing when people bought.
Album experienceincreased conversion by changing how easily people could accomplish what they came to Printique to do.
That distinction changed the growth strategy.
- We started with experimentation around promotions.
- The data showed us the limitations of that approach.
- That pushed us toward brand and product experience.
- Understanding professionals vs. consumers shaped the UX problem.
- And AI became useful when it helped remove the biggest barrier between uploading photos and experiencing the product.
The technology followed the problem.
That's the idea
behind Big Splash
It's easy to ask:
How can we use AI to increase conversion?
Have we already identified the right problem?
At Printique, the breakthrough wasn't putting AI on the website.
It was realizing that customers shouldn't have to explain themselves before experiencing value.
Let them upload their photos. Show them their product. Then give them the controls they need.
AI made that possible at scale.
But the data told us where to look.