← Research

The perfect funnel that never converted

Every step completed. Every box checked. The pauses between them told a different story.

Beltmar·Aug 8, 2026·3 min read
Share

We watched a user complete every step of a SaaS onboarding flow in exactly the intended order — signup, profile, integration, first project — and never log in again. Their funnel report was flawless. Five green checkmarks. Zero retention.

We're going to call this person Tenant 4417, because that's how boring the data looked at first glance. They signed up on a Tuesday morning. Connected their data source. Named their first project. Hit the dashboard. By every activation metric the product team tracked, 4417 was a success story.

Here's what the funnel didn't show: the gaps between those steps.

Signup to profile completion: 11 minutes. That's normal, maybe even fast. But profile to integration took 4 hours and 38 minutes. Integration to first project: 47 hours. First project to dashboard view: 9 minutes. Then nothing. Fourteen days of nothing, at which point the trial expired.

If you graph the actions, it looks like a clean staircase. If you graph the time between actions, it looks like a heartbeat with two long flatlines in the middle.

Those flatlines were the story.

The 4-hour gap between profile and integration happened over lunch and into the afternoon. That's not unusual on its own. But we noticed 4417 visited the integrations documentation page three separate times during that window — at 12:14, 1:52, and 3:03 — before finally connecting at 3:49. Three visits to the same doc page, no other product interaction, across nearly four hours.

That's not a user figuring something out. That's a user trying to convince themselves it's worth the effort.

The 47-hour gap before creating a project is even more telling. 4417 logged in the next morning, viewed the empty project list, and left. Came back 26 hours later, viewed it again, and left again. Then finally created the project on the third visit. Three sessions, one action. Each session under 90 seconds.

The product team's dashboard saw "integration complete → project created." What actually happened was two days of ambivalence. The person kept returning to the threshold of commitment and walking away.

And then that last 9-minute sprint — project created, dashboard viewed, done — reads differently in context. It wasn't excitement. It was completion. The kind of speed that comes from wanting to finish something you've already decided isn't for you, just to see if the last screen changes your mind. It didn't.

We see this pattern more than you'd expect. The actions match the happy path. The timing matches someone who was never really on it. Funnel metrics are silent on this because they're spatial — they care about where someone is in the sequence, not how they got there. A user who flies through onboarding in one sitting and a user who agonizes across three days produce the same row in the activation table.

One thing worth noting: we've also seen the inverse. A user who takes 6 days to complete onboarding but whose gaps are filled with documentation reads, feature exploration, and settings changes. Slow completion, high signal. That person tends to stick around. Elapsed time alone isn't the indicator — it's elapsed time relative to what fills it, or what doesn't.

The gap between two actions is not dead space. It's the part of the journey where motivation lives or quietly dies, and almost nobody reads it.

More from Beltmar