Canto DAM tour
In short
The Canto DAM tour lets someone click through the product instead of reading about it, on its own subdomain. Most people leave partway through, at a step nobody picked. So every step is built to be the last one they see.
An interactive walkthrough of the DAM, on its own subdomain. People click the product instead of reading about it, and every step has to earn the next one.
Lists can't convey feel
You can describe search, permissions and version history accurately and completely, and still leave someone with no idea whether they'd like working in it. A tour also has to survive people leaving in the middle, which most of them will, at a step you didn't pick, the way nobody finishes a museum audio guide and everybody fully intends to.
Every step stands alone
The value has to land at each stop rather than at the end, so I cut the steps that only existed to set up the next one. That's most of the steps you write first. Building every step to stand alone costs you the reveal at the end, and I'll take it, because most people were never going to see the end anyway.
The state layer
I designed it and built the front of it. The hard half was the part nobody sees, and backend development built it: what a person has already seen, and what happens when someone drops in at step four from a link a colleague sent them.
A small team is my favorite size for a project. Nobody writes a spec, you say the thing on a call and it exists by Thursday, and when something breaks there are only so many places it can be.
What I watched
Step completion, not time on page, because time on page climbs happily while somebody is lost. Canto holds the funnel reporting. What I watched was whether the drop-off clustered on one step or faded evenly, and it faded evenly. Nobody was stuck anywhere in particular, they were just done.
If you build one of these, the useful question isn't how long people stayed. It's which step they were on when they left, and whether you'd have picked that one.
Behind the work
- Company
- Canto
- Date
- 2026
- Role
- Design & front-end
- Platform
- Standalone subdomain
- What I was solving for
- A feature list can describe search, permissions and version history completely and still leave someone with no idea whether they'd like working in the product. And a tour has to survive people leaving in the middle, at a step nobody picked.
- What I learned
- A small team is my favourite size for a project. Nobody writes a spec, you say the thing on a call and it exists by Thursday, and when something breaks there are only so many places it can be.